You are here

Home » News Feeds » Sources

The Carbon Brief

Subscribe to The Carbon Brief feed The Carbon Brief
URL: https://www.carbonbrief.org/
Updated: 8 hours 17 min ago

Q&A: Why are UK energy bills so high – and how could they be lowered?

Tue, 10/27/2026 - 09:47
Print article Share Article Contents

Energy bills in the UK have continued to soar in 2026, with intense debate raging about how to bring them down. 

In October 2026, the cap on household energy bills reached its highest level in three years, rising by £60 amid high gas prices driven by the Iran war.

The cap is forecast to jump by as much as another quarter in January 2027, pushing annual energy bills towards £2,200.

Energy bills have not been that high since 2022 when, in the fallout of the Russian invasion of Ukraine, the UK government spent more than £60bn on supporting consumers through the crisis.

While expensive gas is behind the lion’s share of the rise in bills since 2020, policy costs and investment in the gas and electricity grids are also putting some pressure on prices.

This has sparked conflicting calls to continue the clean-energy transition to get off the “fossil-fuel rollercoaster”, to move more slowly – or to abandon climate goals in favour of more gas.

Below, Carbon Brief looks at the real drivers of high energy bills in the UK, what has been done to address them to date and what options are available to cut them going forward.

These include taking more policy costs off electricity bills, boosting flexibility or shifting to more efficient technologies with lower running costs, such as heat pumps and electric vehicles (EVs).

What is driving high energy bills?

In October 2026, UK household energy bills reached their highest levels since 2022, with the price cap for typical homes hitting £1,723 per year, up 47% in real terms since October 2020.

Industrial energy users also face high costs, particularly for their electricity.

For households, another sharp increase is now all but guaranteed under the price cap for January 2027, which is forecast to rise by around 25% from current levels.

Wholesale costs for electricity and gas make up 44% of the current cap of £1,723 per year, according to Carbon Brief analysis of figures from thinktank Nesta, shown in the figure below.

The wholesale cost of both electricity and gas is tied to gas prices, due to the marginal pricing system used in power markets in the UK and most other countries.

Network charges make up a further 24% of bills and operating costs are another 17%, whereas policy costs only account for 7% of the total.

There has been increasing interest in the so-called “non-commodity costs” – those not related to wholesale energy – that now make up a significant portion of bills, as shown in the figure above.

Yet while these costs have been increasing – particularly those related to investing in energy networks – they are not the reason why energy bills have soared once again in recent months.

Indeed, Nesta analysis recently published by Carbon Brief shows that, contrary to widespread UK media and political narratives, a surge in wholesale costs driven by expensive gas is responsible for more than 80% of the rise in dual-fuel household energy bills since 2020, as shown below.

The second-largest driver of the rise in costs since 2020 is due to the cost of operating, maintaining and expanding gas and electricity networks, as shown in the figure above.

At the end of 2025, energy regulator Ofgem approved £29bn of upfront investment into the national gas and electricity grids, as well as the gas distribution network that connects to homes and businesses. This is part of a wider package of up to around £90bn of investment out to 2031.

As a result of these investment plans, network charges are expected to increase by around 50% between 2025 and 2030, from £222 a year to £330 for a typical dual-fuel household.

(This figure covers gas and electricity transmission and gas distribution, but not electricity distribution. A further package of increased investment in the electricity distribution network, for the period from 2028-2033, is due to be approved by Ofgem by the end of 2027.)

#gform_wrapper_7[data-form-index="0"].gform-theme,[data-parent-form="7_0"]{--gf-color-primary: #204ce5;--gf-color-primary-rgb: 32, 76, 229;--gf-color-primary-contrast: #fff;--gf-color-primary-contrast-rgb: 255, 255, 255;--gf-color-primary-darker: #001AB3;--gf-color-primary-lighter: #527EFF;--gf-color-secondary: #fff;--gf-color-secondary-rgb: 255, 255, 255;--gf-color-secondary-contrast: #112337;--gf-color-secondary-contrast-rgb: 17, 35, 55;--gf-color-secondary-darker: #F5F5F5;--gf-color-secondary-lighter: #FFFFFF;--gf-color-out-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-out-ctrl-light-rgb: 17, 35, 55;--gf-color-out-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-out-ctrl-light-lighter: #F5F5F5;--gf-color-out-ctrl-dark: #585e6a;--gf-color-out-ctrl-dark-rgb: 88, 94, 106;--gf-color-out-ctrl-dark-darker: #112337;--gf-color-out-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-color-in-ctrl: #fff;--gf-color-in-ctrl-rgb: 255, 255, 255;--gf-color-in-ctrl-contrast: #112337;--gf-color-in-ctrl-contrast-rgb: 17, 35, 55;--gf-color-in-ctrl-darker: #F5F5F5;--gf-color-in-ctrl-lighter: #FFFFFF;--gf-color-in-ctrl-primary: #204ce5;--gf-color-in-ctrl-primary-rgb: 32, 76, 229;--gf-color-in-ctrl-primary-contrast: #fff;--gf-color-in-ctrl-primary-contrast-rgb: 255, 255, 255;--gf-color-in-ctrl-primary-darker: #001AB3;--gf-color-in-ctrl-primary-lighter: #527EFF;--gf-color-in-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-in-ctrl-light-rgb: 17, 35, 55;--gf-color-in-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-in-ctrl-light-lighter: #F5F5F5;--gf-color-in-ctrl-dark: #585e6a;--gf-color-in-ctrl-dark-rgb: 88, 94, 106;--gf-color-in-ctrl-dark-darker: #112337;--gf-color-in-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-radius: 3px;--gf-font-size-secondary: 14px;--gf-font-size-tertiary: 13px;--gf-icon-ctrl-number: url("data:image/svg+xml,%3Csvg width='8' height='14' viewBox='0 0 8 14' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M4 0C4.26522 5.96046e-08 4.51957 0.105357 4.70711 0.292893L7.70711 3.29289C8.09763 3.68342 8.09763 4.31658 7.70711 4.70711C7.31658 5.09763 6.68342 5.09763 6.29289 4.70711L4 2.41421L1.70711 4.70711C1.31658 5.09763 0.683417 5.09763 0.292893 4.70711C-0.0976311 4.31658 -0.097631 3.68342 0.292893 3.29289L3.29289 0.292893C3.48043 0.105357 3.73478 0 4 0ZM0.292893 9.29289C0.683417 8.90237 1.31658 8.90237 1.70711 9.29289L4 11.5858L6.29289 9.29289C6.68342 8.90237 7.31658 8.90237 7.70711 9.29289C8.09763 9.68342 8.09763 10.3166 7.70711 10.7071L4.70711 13.7071C4.31658 14.0976 3.68342 14.0976 3.29289 13.7071L0.292893 10.7071C-0.0976311 10.3166 -0.0976311 9.68342 0.292893 9.29289Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-select: url("data:image/svg+xml,%3Csvg width='10' height='6' viewBox='0 0 10 6' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M0.292893 0.292893C0.683417 -0.097631 1.31658 -0.097631 1.70711 0.292893L5 3.58579L8.29289 0.292893C8.68342 -0.0976311 9.31658 -0.0976311 9.70711 0.292893C10.0976 0.683417 10.0976 1.31658 9.70711 1.70711L5.70711 5.70711C5.31658 6.09763 4.68342 6.09763 4.29289 5.70711L0.292893 1.70711C-0.0976311 1.31658 -0.0976311 0.683418 0.292893 0.292893Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-search: url("data:image/svg+xml,%3Csvg width='640' height='640' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath d='M256 128c-70.692 0-128 57.308-128 128 0 70.691 57.308 128 128 128 70.691 0 128-57.309 128-128 0-70.692-57.309-128-128-128zM64 256c0-106.039 85.961-192 192-192s192 85.961 192 192c0 41.466-13.146 79.863-35.498 111.248l154.125 154.125c12.496 12.496 12.496 32.758 0 45.254s-32.758 12.496-45.254 0L367.248 412.502C335.862 434.854 297.467 448 256 448c-106.039 0-192-85.962-192-192z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-label-space-y-secondary: var(--gf-label-space-y-md-secondary);--gf-ctrl-border-color: #686e77;--gf-ctrl-size: var(--gf-ctrl-size-md);--gf-ctrl-label-color-primary: #112337;--gf-ctrl-label-color-secondary: #112337;--gf-ctrl-choice-size: var(--gf-ctrl-choice-size-md);--gf-ctrl-checkbox-check-size: var(--gf-ctrl-checkbox-check-size-md);--gf-ctrl-radio-check-size: var(--gf-ctrl-radio-check-size-md);--gf-ctrl-btn-font-size: var(--gf-ctrl-btn-font-size-md);--gf-ctrl-btn-padding-x: var(--gf-ctrl-btn-padding-x-md);--gf-ctrl-btn-size: var(--gf-ctrl-btn-size-md);--gf-ctrl-btn-border-color-secondary: #686e77;--gf-ctrl-file-btn-bg-color-hover: #EBEBEB;--gf-field-img-choice-size: var(--gf-field-img-choice-size-md);--gf-field-img-choice-card-space: var(--gf-field-img-choice-card-space-md);--gf-field-img-choice-check-ind-size: var(--gf-field-img-choice-check-ind-size-md);--gf-field-img-choice-check-ind-icon-size: var(--gf-field-img-choice-check-ind-icon-size-md);--gf-field-pg-steps-number-color: rgba(17, 35, 55, 0.8);}

CommentsThis field is for validation purposes and should be left unchanged.

Free weekly newsletter

.preheader p{ margin-top: 0; font-family: 'PT Sans', sans-serif; font-weight: var(--type--3--font-weight--bold); color: var(--button--color); font-size: var(--button--font-size, inherit); } .newsletter-inline{ display: flex; border: solid 1px #333333; padding: 1em; background: #ffffff; } .inline-email{ display:inline-block; margin-right:1em; margin-top:0 !important; margin-bottom:0.5em; } #field_submit{ display:inline-block; margin-top:0 !important; } .gform_wrapper .gfield+.gfield{ margin-top:0 } Email gform.initializeOnLoaded( function() {gformInitSpinner( 7, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery('#gform_ajax_frame_7').on('load',function(){var contents = jQuery(this).contents().find('*').html();var is_postback = contents.indexOf('GF_AJAX_POSTBACK') >= 0;if(!is_postback){return;}var form_content = jQuery(this).contents().find('#gform_wrapper_7');var is_confirmation = jQuery(this).contents().find('#gform_confirmation_wrapper_7').length > 0;var is_redirect = contents.indexOf('gformRedirect(){') >= 0;var is_form = form_content.length > 0 && ! is_redirect && ! is_confirmation;var mt = parseInt(jQuery('html').css('margin-top'), 10) + parseInt(jQuery('body').css('margin-top'), 10) + 100;if(is_form){jQuery('#gform_wrapper_7').html(form_content.html());if(form_content.hasClass('gform_validation_error')){jQuery('#gform_wrapper_7').addClass('gform_validation_error');} else {jQuery('#gform_wrapper_7').removeClass('gform_validation_error');}setTimeout( function() { /* delay the scroll by 50 milliseconds to fix a bug in chrome */ jQuery(document).scrollTop(jQuery('#gform_wrapper_7').offset().top - mt); }, 50 );if(window['gformInitDatepicker']) {gformInitDatepicker();}if(window['gformInitPriceFields']) {gformInitPriceFields();}var current_page = jQuery('#gform_source_page_number_7').val();gformInitSpinner( 7, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery(document).trigger('gform_page_loaded', [7, current_page]);window['gf_submitting_7'] = false;}else if(!is_redirect){var confirmation_content = jQuery(this).contents().find('.GF_AJAX_POSTBACK').html();if(!confirmation_content){confirmation_content = contents;}jQuery('#gform_wrapper_7').replaceWith(confirmation_content);jQuery(document).scrollTop(jQuery('#gf_7').offset().top - mt);jQuery(document).trigger('gform_confirmation_loaded', [7]);window['gf_submitting_7'] = false;wp.a11y.speak(jQuery('#gform_confirmation_message_7').text());}else{jQuery('#gform_7').append(contents);if(window['gformRedirect']) {gformRedirect();}}jQuery(document).trigger("gform_pre_post_render", [{ formId: "7", currentPage: "current_page", abort: function() { this.preventDefault(); } }]); if (event && event.defaultPrevented) { return; } const gformWrapperDiv = document.getElementById( "gform_wrapper_7" ); if ( gformWrapperDiv ) { const visibilitySpan = document.createElement( "span" ); visibilitySpan.id = "gform_visibility_test_7"; gformWrapperDiv.insertAdjacentElement( "afterend", visibilitySpan ); } const visibilityTestDiv = document.getElementById( "gform_visibility_test_7" ); let postRenderFired = false; function triggerPostRender() { if ( postRenderFired ) { return; } postRenderFired = true; gform.core.triggerPostRenderEvents( 7, current_page ); if ( visibilityTestDiv ) { visibilityTestDiv.parentNode.removeChild( visibilityTestDiv ); } } function debounce( func, wait, immediate ) { var timeout; return function() { var context = this, args = arguments; var later = function() { timeout = null; if ( !immediate ) func.apply( context, args ); }; var callNow = immediate && !timeout; clearTimeout( timeout ); timeout = setTimeout( later, wait ); if ( callNow ) func.apply( context, args ); }; } const debouncedTriggerPostRender = debounce( function() { triggerPostRender(); }, 200 ); if ( visibilityTestDiv && visibilityTestDiv.offsetParent === null ) { const observer = new MutationObserver( ( mutations ) => { mutations.forEach( ( mutation ) => { if ( mutation.type === 'attributes' && visibilityTestDiv.offsetParent !== null ) { debouncedTriggerPostRender(); observer.disconnect(); } }); }); observer.observe( document.body, { attributes: true, childList: false, subtree: true, attributeFilter: [ 'style', 'class' ], }); } else { triggerPostRender(); } } );} );

The increase in network costs is often attributed to the need to connect increasing levels of geographically dispersed renewable energy sources, particularly wind power.

However, a significant proportion of these increases relate to the need to update ageing infrastructure after decades of under-investment.

This package of investment represents a “once-in-a-generation renewal of our energy infrastructure”, Ofgem head Tim Jarvis told a recent conference attended by Carbon Brief.

Ofgem says some three-fifths of the £108 per household per year rise in network charges to 2030 is due to “business-as-usual” costs such as ongoing maintenance of the gas and electricity networks. It says only £30 is due to investing in expanding the electricity transmission grid.

Similarly, according to the Energy Networks Association, up to 80% of the investment in electricity grids is required to maintain existing assets, with the remainder going towards expansion.

Another point of contention is “constraint costs”, where windfarms are switched off if the network lacks the capacity to transmit their power – and gas plants are switched on to meet demand.

These costs sit within network balancing charges and – despite rising steeply in recent years, partly due to high gas prices – these still make up just £35, or 4% of typical annual electricity bills.

Today, rising constraint costs are partly a reflection of the decision in the 2010s to postpone network investments under a policy known as “connect and manage”. Ofgem says the investments it has now approved in the electricity grid would help limit constraints in the future.

Trade association RenewableUK’s head of policy Steven Agnew points to a recent review by the National Audit Office and says investment in the grid is “critical to avoid large rises in consumers’ bills as more of the country’s electricity is generated from clean sources”. He adds: 

“The current crunch is the result of inaction by successive previous governments to tackle this issue. Consumers are losing out on some of the benefits because low-cost wind has to be turned down and expensive gas turned up elsewhere.”

Projections from the National Energy System Operator (Neso) illustrate how constraint costs depend on the relative pace of expansion in new generating capacity and the electricity network.

It says balancing costs could rise from £2bn to £8bn by 2030, if the grid faces delays. However, its central projection is £3-4bn, with a range of options available to limit this increase.

Finally, following years of elevated energy bills, the cost of “bad debt” has increased. This is where unpaid bills are recovered by energy companies through a charge on all households.

These charges add around £60 to typical bills, according to Carbon Brief analysis of Ofgem data. This is an increase of around £40 since 2020 – more than half of the rise in “operating costs”. Trade association Energy UK says rising debt could add another £10-15 to bills by next year.

What has been done to limit bills so far?

The Labour government was elected in 2024 on a manifesto that pledged to cut bills by £300 per year by 2030, relative to levels in late 2023.

The government has taken a number of steps in this direction.

At the 2025 budget, it closed the “energy company obligation” (ECO) efficiency scheme and moved 75% of the “renewables obligation” support scheme for legacy renewables into general taxation.

Collectively, these two changes were expected to save a typical household £134 per year, or £154 when including households that use electric heating. (These savings only apply to households.)

Many groups, from industry to politicians, had long called for policy costs to be taken off electricity bills in an effort to bring down prices, but this had previously been resisted by the Treasury.

In addition, new prime minister Andy Burnham announced in July that his government would remove VAT from electricity bills for six months from 1 October, cutting bills by around £45.

Yet despite these efforts, bills have continued to increase.

Speaking to Carbon Brief, Dr Richard Carmichael, a researcher at the UK Energy Research Centre (UKERC) and research fellow at Imperial College London, says the impact of reducing policy costs and VAT has been “masked” by rising wholesale energy prices.

As a result, there are calls for the government to go even further in its efforts to reduce policy costs – or to take more radical steps to cut bills. (See: What are the options to bring down bills?)

One factor that is less well-appreciated is the impact that clean-power sources have already had on energy bills. Commentators tend to focus narrowly on the cost of support schemes for these projects, without considering the impact they have on the total cost of running the energy system.

Nesta research suggests that electricity bills are rising “much more slowly” than gas this year, despite the impact of the Hormuz crisis, due to the expansion of clean power. It says this is an early sign of electricity prices starting to “decouple” from gas prices.

In April, the government announced reforms that it hopes will help “break [the] link between gas and electricity prices, but analysts told Carbon Brief the impact was likely to be “modest”.

Nevertheless, the link between the two will continue to weaken as more wind and solar gets built, including projects supported by a record “contracts for difference” (CfD) auction earlier this year.

Since the start of the Iran war, the UK has avoided the need to import gas worth around £5.9bn as a result of record wind and solar output, according to previous Carbon Brief analysis.

RenewableUK’s Agnew tells Carbon Brief: “Put simply, every wind turbine we build is gas we don’t need to buy.”

In addition to directly displacing gas, the windfarms on the system cut wholesale power prices by around a third in 2025, according to the Energy and Climate Intelligence Unit (ECIU).

This is due to the “merit-order effect”, whereby windfarms push the most expensive gas plants off the system, reducing the average wholesale power price.

According to separate research published by University College London, UK wind energy cut consumer costs by a total of £104bn over 2010-2023, net of subsidy costs, through a combination of the merit-order effect and reduced gas demand, which lowered gas prices.

What options are there for bringing down bills?

With household and industrial prices remaining elevated for more than five years, countless ideas have been presented for bringing down the cost of energy.

These span from immediate but more modest tweaks that could have a rapid impact, through to longer-term changes in the energy system, the design of energy markets or retail tariffs.

Moreover, amid a collapse in the UK consensus, some politicians want climate goals to be abandoned in the name of cheaper energy. (See: Is there a ‘cheap power’ alternative if the UK slows down?)

Ahead of the budget on 28 October, the most widely shared request is for the chancellor to take more policy costs off electricity bills, either for households only, or for all energy users.

A September letter to the government from more than 120 trade bodies, businesses and charities calls for “hidden taxes” to be taken off energy bills for both businesses and households.

.cb-tweet img{ border: solid 1.25px #333333; border-radius: 5px; } @media (max-width:650px){ .cb-tweet{ width:100%; } }

It argues that a series of levies should be paid out of general taxation, in addition to the existing decision to take 75% of renewables obligation costs off household electricity bills.

This would include a levy supporting the Sizewell C new nuclear plant, the cost of the warm home discount scheme for fuel-poor households and feed-in tariffs for small-scale legacy renewables.

The group says that removing these additional policy costs from bills would “bring the total average domestic energy bill reduction achieved by this government to as much as £250 a year”.

It adds that this would bring down inflation, boost manufacturing and incentivise investment in the electrification of the economy, helping to “fully realise [the UK’s] economic potential”.

Speaking to Carbon Brief, energy consultant Kate Mulvany says bringing policy costs and bad debt into general taxation is “in all likelihood the fastest route” to reducing bills.

The main barrier to removing further policy and debt costs from bills is that the money would have to come from somewhere else – most likely the Treasury, which has resisted this in the past.

The decision to take 75% of the renewables obligation off household bills is seen as a breakthrough, with Energy UK head Dhara Vyas saying the Treasury should go a “step further”.

Alternatively, a report from parliament’s Energy Security and Net Zero Committee calls for policy costs to be moved from electricity onto gas bills.

This would help “rebalance” the cost of electricity relative to gas, encouraging consumers to switch to more efficient electric technologies, such as heat pumps and electric cars.

This is because it would reduce the “spark gap”, the ratio between electricity and gas prices. This ratio stood at 4.3 in the UK as of April 2026, although it has dropped to 3.3 under the latest price cap and could fall again to 3.0 in January 2027, according to Carbon Brief analysis.

Further options to bring down costs for consumers have been suggested by groups including Nesta, the “electricity bills taskforce” and many others.

Ideas include the government paying off household energy debt, instead of recovering it through bills, a permanent removal of VAT on electricity prices and boosting electricity system “flexibility” so as to maximise the use of existing network assets.

Carmichael tells Carbon Brief:

“Flexibility is a fairly obvious one that…would deliver bill reductions for people who can participate in flexibility, but it would also reduce system costs for everyone.”

Ashutosh Padelkar, research lead at Aurora, agrees that there is “more work to be done on improving network efficiency”. In addition to boosting flexibility, he points to innovations such as “dynamic line ratings”, where cables can be “overloaded a bit for a small amount of time”.

With bills set to leap by around a quarter at the start of 2027, there have also been calls for specific measures to offer households relief at the budget.

Many argue that this support should be targeted to limit costs, for example through extending the “warm homes discount”. However, Mulvany warns that this could increase costs for other households if it is paid for via energy bills, as is currently the case for the warm homes discount.

Another option is to introduce a “social tariff” for fuel-poor households. Dr Carmichael notes that while this has some positives, it also has “potential for a sort of a cliff edge where people are just missing out on it”. This can make them challenging to put together in an equitable way. 

Meanwhile, the New Economics Foundation thinktank – where energy secretary Miatta Fahnbulleh used to work – suggests a “rising block tariff”, where consumers pay lower rates for a set amount of “essential” energy use. However, there are fears this could discourage electrification.

Ahead of the budget, Burnham has asked ministers to look at developing a plan to remove standing charges from gas bills. The idea was put forward by Nesta, which says it would save typical households £27 per year, while helping encourage the shift to heat pumps and EVs.

Finally, separate proposals from thinktank Common Wealth and consultancy Stonehaven with Greenpeace, call for more radical changes that would see gas-fired power stations taken into a strategic reserve, in order to more completely break the link between gas and power prices.

Is there a ‘cheap power’ alternative if the UK slows down?

Some groups argue that the UK needs a fundamental shift in its energy strategy in order to tackle high bills. Such calls are often based on the idea that the government is “rushing” to meet its target for clean power by 2030 – and that slowing down would cut costs.

For example, a September 2026 report from the Institute for Fiscal Studies (IFS) states that “continuing to push for rapid decarbonisation does risk increasing costs”.

Its argument rests on NESO modelling, published in 2024, which suggests it could be slightly cheaper to rely on more gas in 2030 than planned under the clean-power 2030 target.

Peter Levell, IFS deputy research director, tells Carbon Brief that gas prices have “risen significantly” since the NESO modelling was published, which “other things equal” would make it relatively cheaper to push for more renewables instead of relying on more gas power.

However, he adds that capital costs have also risen, arguing that this makes renewables relatively more expensive. (The cost of gas-fired power plants has risen even more steeply.) Levell adds that it is unclear how long the current disruption causing high gas prices will last.

Another prominent – and repeated – voice for slowing down the government’s push for clean power has come from the Tony Blair Institute (TBI).

In a February 2026 report, for example, the TBI argues that the government should set an objective of “cheaper power by 2030” instead of “clean power 2030”. It says this could be done “while maintaining a clear decarbonisation trajectory”.

In earlier versions of this argument, the TBI called for a slower expansion of renewables, in favour of greater reliance on nuclear power and carbon-capture and storage.

However, perhaps implicitly acknowledging that these are not cheap technologies, its latest report focuses more on calls for flexibility, optimised grid expansion and a push for electrification.

Perhaps the strongest push to abandon the government’s current approach has come from the opposition Conservative party, which now lobbies hard against clean power 2030, in addition to calling for the UK to scrap all of its climate targets and the legislation that underpins them. 

The party backed a report by thinktank Onward, which claimed that the UK could save “over £320bn” with a push for “cheap power” instead of “clean power”. However, a Carbon Brief factcheck found it relied on a series of questionable assumptions.

Ultimately, the report built its conclusions in from the start by assuming that gas and nuclear power would be cheap. As Tom Glover, head of energy firm RWE UK, put it at the time:

“There’s a difference between a useful scenario and a plan you’d build a system around. This report is the former presented as the latter. If you accept its assumptions, the results follow logically. The problem is the assumptions…built on a stack of optimistic inputs.”

Similarly, Johnny Gowdy, a director at thinktank Regen, argues in a paper responding to the Onward report that its gas-price assumptions are “very optimistic – and potentially reckless”.

Following the start of the Iran war, UK wholesale gas prices – which were already roughly three-times higher than before the Ukraine crisis – have more than doubled again.

According to RenewableUK, the UK is currently experiencing its ninth gas-price “spike” of the past decade. Mulvany tells Carbon Brief that such spikes are “normal”. She adds:

“Wars are fought over energy and resources – and this is the way things have always been…[T]his isn’t a blip. If you look historically, this is normal.”

(The total cost to the UK economy of higher energy prices in the wake of Russia’s invasion of Ukraine amounts to £183bn over four years, according to the ECIU.)

NESO notes that under pathways with a larger role for gas plants, the nation’s energy system would remain “more exposed to volatile international gas prices”.

As such, the UK’s transition is about finding the right balance between the up-front cost of investment in clean-energy alternatives and the ongoing risk of exposure to volatile gas prices.

The IFS also notes that there are additional positives from moving away from fossil fuels, including improved local air quality, less price volatility and potentially greater security of energy supply.

Citing Carbon Brief analysis published in 2024, RenewableUK’s Agnew argues that the government should learn from past “mistakes”.

He points to the then-Conservative government decision from 2015 to “cut the green crap”, which ultimately added £22bn to bills by leaving the UK “more exposed to volatile fossil-fuel prices”.

UKERC’s Carmichael echoes these points, telling Carbon Brief: 

“If we’re not transitioning to clean power, which for the UK would mean more independence and more energy sovereignty, I don’t think there is an alternative…[and] if we don’t do it, then we’re still going to be vulnerable to volatility and fossil-fuel prices.”

Finally, many of the arguments against the clean power 2030 target are based on the idea that it will be extremely difficult to meet and that closing the gap would be costly.

Indeed, recent analysis by consultancy LCP Delta suggests that by 2030, the UK will only reach 83% clean power, with the 95% target unlikely to be hit until 2035.

Yet this reality is implicitly acknowledged by the government, which has said that it will not contract for more clean power towards the 2030 target unless it delivers lower bills.

Speaking at the Energy UK conference in October, Fahnbulleh said:

“We all know that one of the reasons we are seeing rising prices is because we are uniquely exposed to global fossil-fuel markets…There will be people that will say you should slow down…But I believe that we must seize the opportunity of this moment.”

Can electrification of homes and businesses help cut bills?

While there is political disagreement over the best mix of power supplies, there is far greater agreement on the need to expand the use electric technologies such as EVs and heat pumps.

For example, while the Conservative party opposes the government’s clean power plan, it argues that the UK needs to electrify its economy to ensure both “prosperity and a better environment”.

At the Energy UK conference in October, delegates listed “electrification” and “affordability” as the top-two energy related priorities for the government.

There is significant overlap between these two goals.

At present, most households pay separate bills for heating, power and transport. This means they are paying to run three separate energy systems for gas, electricity and petrol.

If the UK largely electrifies its economy, households will instead rely on a single system that heats and powers their homes, as well as fuelling their cars. By boosting consumption, electrification also spreads the fixed costs of upgrading the network across a larger base of demand.

Padelkar tells Carbon Brief that there “isn’t enough electrification in the system” at the moment:

“A significant amount of heating demand is gas-fired…[I]f that demand were to sit on the electricity side instead, it would allow for costs to be shared more fairly across a greater base of consumers.”

According to energy firm SSE, the shift from three “parallel” energy systems to a single one based on electricity would reduce consumer costs over time, as shown in the figure below.

Credit: SSE

In this context, the usual focus on “dual-fuel” household gas and electricity bills, in isolation, misses the potential for savings due to consolidating not only heat and power, but also car use.

Indeed, electric cars are currently up to nine times cheaper to run than petrol cars in the UK, according to Carbon Brief analysis.

The savings from electrification are partly a result of the far greater efficiency of EVs and heat pumps, relative to combustion cars and gas boilers. Reduced fossil-fuel imports also play a role.

Very similar arguments have been put forward by the UK government’s advisory Climate Change Committee (CCC), the International Energy Agency and others.

A recent CCC report, for example, said that faster electrification “would cut UK household bills”.

It outlined how electrification would not only reduce the cost of the UK energy system in the future, but is already able to cut household costs today, as shown in the figure below. Crucially, its analysis included the higher upfront costs of investing in electric technologies, such as heat pumps.

Household energy costs for heat, power and transport, £ per year. The upfront costs of purchasing cars, heating systems, chargers and solar panels are annualised. Source: CCC progress report 2026.

The government also supports this argument, having set out a £13bn “warm homes plan” that it said could cut household bills by up to £600 per year through investment in energy efficiency measures and clean technologies, such as heat pumps, solar panels and batteries.

Vyas told the Energy UK conference earlier this month: 

“The industry is clear that a system largely supplied by clean power and an electrified economy represents the way forward. This is not driven by ideology but by evidence.”

Related UK budget 2026: Key climate and energy announcements 28.10.2026 UK policy Analysis: How the UK spent more than £11.6bn on climate aid since 2021 28.10.2026 UK policy Revealed: UK relied on £2bn of ‘questionable accounting’ to hit its climate-aid goal 28.10.2026 UK policy Revealed: Expensive gas behind ‘nearly 90%’ of rise in UK electricity bills 21.10.2026 Oil and gas

The post Q&A: Why are UK energy bills so high – and how could they be lowered? appeared first on Carbon Brief.

Categories: I. Climate Science

Guest post: Greenland ice sheet shrinks for 30th year in a row

Tue, 10/27/2026 - 07:07
Print article Share Article Contents

This guest post is by:

Dr Martin Stendel, climate scientist at the Danish Meteorological Institute (DMI), part of the Polar Portal 

Prof Ruth Mottram, climate scientist at DMI and honorary professor at Aarhus University

Greenland has now marked the 30th consecutive year where it lost more ice than it gained. 

Over the 12-month period to August 2026, the world’s second-largest ice sheet lost around 135bn tonnes of ice. 

This major ice loss came as the region experienced a relatively cool start to summer and less surface melt than in previous years. 

In other words, it was not an “extreme” year for weather in Greenland – and, yet, the ice sheet ended firmly in the red.

Three decades of continuous ice loss in Greenland is significant, given that scientists typically need 30 years of data to separate the “fingerprint” of human-caused climate change from natural variations in the climate. 

In Greenland, the direction of travel has been the same – downward – every single year for 30 years.

These latest figures come as a new report highlights that ice melting processes seen for decades in Greenland are increasingly being replicated in Antarctica.

Greenland, therefore, is not only the largest present-day source of ice-driven sea level rise – but offers a preview of what is in store for the world’s largest ice sheet. 

(For our previous analyses of Greenland’s ice cover, see coverage in 2025, 2024, 2023, 2022, 2021, 2020, 2019, 2018, 2017, 2016 and 2015.)

Surface mass balance

One way that scientists track the health of the Greenland ice sheet is by tracking changes to its “surface mass balance”.

This is the difference between ice gains from snowfall and losses from melt at the surface of Greenland’s ice sheet.

Greenland’s annual ice cycle runs from 1 September through to the end of August. The surface mass balance at the end of this period has, to date, always been positive, because the surface accumulates more snow than it loses from melting ice.  

The ice sheet typically gains mass during the colder months of autumn, winter and spring. Then, as temperatures increase from around mid-June, the ice sheet begins to lose more ice through surface melt than it gains from snow. 

(Greenland also loses ice when icebergs break off – or “calve” – into the sea, as well when the front of glaciers melt when they meet warm sea water. These losses are included in the “total mass balance” – see below.)

Greenland ended 2025-26 with an overall surface mass balance of 519bn tonnes.

This is the sixth highest in a dataset stretching back 46 years and about 178bn tonnes above the 1981-2010 average of roughly 350bn tonnes.

The map below shows the difference between the surface mass balance in 2025-26 and the 1981-2010 average, where blue shows ice gain and red shows ice loss.

Map showing the difference between the annual SMB in 2025-26 and the 1981-2010 period in mm of ice melt. Blue shows ice gain compared to average and red shows ice loss. Credit: Polar Portal.

The blue lines in the charts below show the surface mass balance throughout 2025-26 for individual days (top) and cumulatively throughout the year (bottom). They highlight how Greenland gained ice between September 2025 and May 2026, before starting to lose ice from June onwards.

Daily (upper) and cumulative (lower) surface mass balance (SMB) of the Greenland ice sheet, in billions of tonnes. Blue lines show the 2025-26 surface mass balance year; grey lines and shading show the 1981-2010 average and variability. Red line shows the record-low year of 2011-12. Credit: Polar Portal.
Total mass balance

Another way scientists measure the health of the Greenland ice sheet is by looking at its “total mass balance”.

This calculation accounts for ice loss at the surface, as well as the calving of icebergs and the melting of glacier fronts where they meet warm seawater. The latter two components are known as the “marine mass balance”.

It also accounts for a small amount of ice melt at the underside of the ice sheet. This is known as the “basal mass balance”.

In recent decades, Greenland’s total mass balance at the end of the ice cycle year has been negative.

The 12 months to August 2026 were no different. Once all these factors are considered, Greenland lost around 135bn tonnes of ice.

We reached this figure using the same method featured in previous Carbon Brief analyses of Greenland ice cover. This approach is based on an average of three climate models. 

#gform_wrapper_5[data-form-index="0"].gform-theme,[data-parent-form="5_0"]{--gf-color-primary: #204ce5;--gf-color-primary-rgb: 32, 76, 229;--gf-color-primary-contrast: #fff;--gf-color-primary-contrast-rgb: 255, 255, 255;--gf-color-primary-darker: #001AB3;--gf-color-primary-lighter: #527EFF;--gf-color-secondary: #fff;--gf-color-secondary-rgb: 255, 255, 255;--gf-color-secondary-contrast: #112337;--gf-color-secondary-contrast-rgb: 17, 35, 55;--gf-color-secondary-darker: #F5F5F5;--gf-color-secondary-lighter: #FFFFFF;--gf-color-out-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-out-ctrl-light-rgb: 17, 35, 55;--gf-color-out-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-out-ctrl-light-lighter: #F5F5F5;--gf-color-out-ctrl-dark: #585e6a;--gf-color-out-ctrl-dark-rgb: 88, 94, 106;--gf-color-out-ctrl-dark-darker: #112337;--gf-color-out-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-color-in-ctrl: #fff;--gf-color-in-ctrl-rgb: 255, 255, 255;--gf-color-in-ctrl-contrast: #112337;--gf-color-in-ctrl-contrast-rgb: 17, 35, 55;--gf-color-in-ctrl-darker: #F5F5F5;--gf-color-in-ctrl-lighter: #FFFFFF;--gf-color-in-ctrl-primary: #204ce5;--gf-color-in-ctrl-primary-rgb: 32, 76, 229;--gf-color-in-ctrl-primary-contrast: #fff;--gf-color-in-ctrl-primary-contrast-rgb: 255, 255, 255;--gf-color-in-ctrl-primary-darker: #001AB3;--gf-color-in-ctrl-primary-lighter: #527EFF;--gf-color-in-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-in-ctrl-light-rgb: 17, 35, 55;--gf-color-in-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-in-ctrl-light-lighter: #F5F5F5;--gf-color-in-ctrl-dark: #585e6a;--gf-color-in-ctrl-dark-rgb: 88, 94, 106;--gf-color-in-ctrl-dark-darker: #112337;--gf-color-in-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-radius: 3px;--gf-font-size-secondary: 14px;--gf-font-size-tertiary: 13px;--gf-icon-ctrl-number: url("data:image/svg+xml,%3Csvg width='8' height='14' viewBox='0 0 8 14' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M4 0C4.26522 5.96046e-08 4.51957 0.105357 4.70711 0.292893L7.70711 3.29289C8.09763 3.68342 8.09763 4.31658 7.70711 4.70711C7.31658 5.09763 6.68342 5.09763 6.29289 4.70711L4 2.41421L1.70711 4.70711C1.31658 5.09763 0.683417 5.09763 0.292893 4.70711C-0.0976311 4.31658 -0.097631 3.68342 0.292893 3.29289L3.29289 0.292893C3.48043 0.105357 3.73478 0 4 0ZM0.292893 9.29289C0.683417 8.90237 1.31658 8.90237 1.70711 9.29289L4 11.5858L6.29289 9.29289C6.68342 8.90237 7.31658 8.90237 7.70711 9.29289C8.09763 9.68342 8.09763 10.3166 7.70711 10.7071L4.70711 13.7071C4.31658 14.0976 3.68342 14.0976 3.29289 13.7071L0.292893 10.7071C-0.0976311 10.3166 -0.0976311 9.68342 0.292893 9.29289Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-select: url("data:image/svg+xml,%3Csvg width='10' height='6' viewBox='0 0 10 6' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M0.292893 0.292893C0.683417 -0.097631 1.31658 -0.097631 1.70711 0.292893L5 3.58579L8.29289 0.292893C8.68342 -0.0976311 9.31658 -0.0976311 9.70711 0.292893C10.0976 0.683417 10.0976 1.31658 9.70711 1.70711L5.70711 5.70711C5.31658 6.09763 4.68342 6.09763 4.29289 5.70711L0.292893 1.70711C-0.0976311 1.31658 -0.0976311 0.683418 0.292893 0.292893Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-search: url("data:image/svg+xml,%3Csvg width='640' height='640' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath d='M256 128c-70.692 0-128 57.308-128 128 0 70.691 57.308 128 128 128 70.691 0 128-57.309 128-128 0-70.692-57.309-128-128-128zM64 256c0-106.039 85.961-192 192-192s192 85.961 192 192c0 41.466-13.146 79.863-35.498 111.248l154.125 154.125c12.496 12.496 12.496 32.758 0 45.254s-32.758 12.496-45.254 0L367.248 412.502C335.862 434.854 297.467 448 256 448c-106.039 0-192-85.962-192-192z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-label-space-y-secondary: var(--gf-label-space-y-md-secondary);--gf-ctrl-border-color: #686e77;--gf-ctrl-size: var(--gf-ctrl-size-md);--gf-ctrl-label-color-primary: #112337;--gf-ctrl-label-color-secondary: #112337;--gf-ctrl-choice-size: var(--gf-ctrl-choice-size-md);--gf-ctrl-checkbox-check-size: var(--gf-ctrl-checkbox-check-size-md);--gf-ctrl-radio-check-size: var(--gf-ctrl-radio-check-size-md);--gf-ctrl-btn-font-size: var(--gf-ctrl-btn-font-size-md);--gf-ctrl-btn-padding-x: var(--gf-ctrl-btn-padding-x-md);--gf-ctrl-btn-size: var(--gf-ctrl-btn-size-md);--gf-ctrl-btn-border-color-secondary: #686e77;--gf-ctrl-file-btn-bg-color-hover: #EBEBEB;--gf-field-img-choice-size: var(--gf-field-img-choice-size-md);--gf-field-img-choice-card-space: var(--gf-field-img-choice-card-space-md);--gf-field-img-choice-check-ind-size: var(--gf-field-img-choice-check-ind-size-md);--gf-field-img-choice-check-ind-icon-size: var(--gf-field-img-choice-check-ind-icon-size-md);--gf-field-pg-steps-number-color: rgba(17, 35, 55, 0.8);}

FacebookThis field is for validation purposes and should be left unchanged.

Free fortnightly newsletter

.preheader p{ margin-top: 0; font-family: 'PT Sans', sans-serif; font-weight: var(--type--3--font-weight--bold); color: var(--button--color); font-size: var(--button--font-size, inherit); } .newsletter-inline{ display: flex; border: solid 1px #333333; padding: 1em; background: #ffffff; } .inline-email{ display:inline-block; margin-right:1em; margin-top:0 !important; margin-bottom:0.5em; } #field_submit{ display:inline-block; margin-top:0 !important; } .gform_wrapper .gfield+.gfield{ margin-top:0 } Email gform.initializeOnLoaded( function() {gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery('#gform_ajax_frame_5').on('load',function(){var contents = jQuery(this).contents().find('*').html();var is_postback = contents.indexOf('GF_AJAX_POSTBACK') >= 0;if(!is_postback){return;}var form_content = jQuery(this).contents().find('#gform_wrapper_5');var is_confirmation = jQuery(this).contents().find('#gform_confirmation_wrapper_5').length > 0;var is_redirect = contents.indexOf('gformRedirect(){') >= 0;var is_form = form_content.length > 0 && ! is_redirect && ! is_confirmation;var mt = parseInt(jQuery('html').css('margin-top'), 10) + parseInt(jQuery('body').css('margin-top'), 10) + 100;if(is_form){jQuery('#gform_wrapper_5').html(form_content.html());if(form_content.hasClass('gform_validation_error')){jQuery('#gform_wrapper_5').addClass('gform_validation_error');} else {jQuery('#gform_wrapper_5').removeClass('gform_validation_error');}setTimeout( function() { /* delay the scroll by 50 milliseconds to fix a bug in chrome */ jQuery(document).scrollTop(jQuery('#gform_wrapper_5').offset().top - mt); }, 50 );if(window['gformInitDatepicker']) {gformInitDatepicker();}if(window['gformInitPriceFields']) {gformInitPriceFields();}var current_page = jQuery('#gform_source_page_number_5').val();gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery(document).trigger('gform_page_loaded', [5, current_page]);window['gf_submitting_5'] = false;}else if(!is_redirect){var confirmation_content = jQuery(this).contents().find('.GF_AJAX_POSTBACK').html();if(!confirmation_content){confirmation_content = contents;}jQuery('#gform_wrapper_5').replaceWith(confirmation_content);jQuery(document).scrollTop(jQuery('#gf_5').offset().top - mt);jQuery(document).trigger('gform_confirmation_loaded', [5]);window['gf_submitting_5'] = false;wp.a11y.speak(jQuery('#gform_confirmation_message_5').text());}else{jQuery('#gform_5').append(contents);if(window['gformRedirect']) {gformRedirect();}}jQuery(document).trigger("gform_pre_post_render", [{ formId: "5", currentPage: "current_page", abort: function() { this.preventDefault(); } }]); if (event && event.defaultPrevented) { return; } const gformWrapperDiv = document.getElementById( "gform_wrapper_5" ); if ( gformWrapperDiv ) { const visibilitySpan = document.createElement( "span" ); visibilitySpan.id = "gform_visibility_test_5"; gformWrapperDiv.insertAdjacentElement( "afterend", visibilitySpan ); } const visibilityTestDiv = document.getElementById( "gform_visibility_test_5" ); let postRenderFired = false; function triggerPostRender() { if ( postRenderFired ) { return; } postRenderFired = true; gform.core.triggerPostRenderEvents( 5, current_page ); if ( visibilityTestDiv ) { visibilityTestDiv.parentNode.removeChild( visibilityTestDiv ); } } function debounce( func, wait, immediate ) { var timeout; return function() { var context = this, args = arguments; var later = function() { timeout = null; if ( !immediate ) func.apply( context, args ); }; var callNow = immediate && !timeout; clearTimeout( timeout ); timeout = setTimeout( later, wait ); if ( callNow ) func.apply( context, args ); }; } const debouncedTriggerPostRender = debounce( function() { triggerPostRender(); }, 200 ); if ( visibilityTestDiv && visibilityTestDiv.offsetParent === null ) { const observer = new MutationObserver( ( mutations ) => { mutations.forEach( ( mutation ) => { if ( mutation.type === 'attributes' && visibilityTestDiv.offsetParent !== null ) { debouncedTriggerPostRender(); observer.disconnect(); } }); }); observer.observe( document.body, { attributes: true, childList: false, subtree: true, attributeFilter: [ 'style', 'class' ], }); } else { triggerPostRender(); } } );} );

However, for the first time, this year’s update includes a separate estimate for Greenland ice loss, based on data from one climate model and automatic weather stations, satellites, GPS data and observations from planes.

This approach estimates that the ice sheet lost an even greater 173bn tonnes of ice over 2025-26.

(Scientific estimates for ice loss in Greenland vary because there is incomplete information about how thick the ice sheet is, how fast it is moving as well as the rate at which it is melting at the surface. Scientists use different approaches for accounting for these uncertainties.)

The chart below shows how Greenland’s total mass balance has changed since 1985, featuring the higher 173bn estimate for the last ice cycle year. 

It shows how 2025-26 was the 30th consecutive year of ice loss in Greenland, with 1996 the last year that saw an annual net gain of ice.

Labels on the x-axis denote the period 1 September to 31 August, where 1987 covers 1 September 1986 to 31 August 1987.

The chart below shows how Greenland’s surface mass balance, marine mass balance and basal mass balance have changed since 1985. All of these components are combined to calculate the ice sheet’s total mass balance (red). 

Labels on the x-axis denote the period 1 September to 31 August, where 1987 covers 1 September 1986 to 31 August 1987.
Surface melt

Winter and early spring in 2025-26 were unusually dry in the southern half of Greenland, leaving around 40bn tonnes less snow on the ice sheet than normal. 

Dry winters typically mean high ice melt rates later on, given that heavier winter snowfall often acts as a protective shield that prevents ice loss in the early part of summer.

However, the relatively dry winter was followed by a relatively cool start of summer. Frequent snowfall and extensive cloud cover limited the sun’s ability to warm the surface at the time of year when solar radiation is typically strongest. 

The onset of the melt season – the first of at least three consecutive days with melting over more than 5% of the ice sheet – occurred on 5 June, nine days later than the 1981-2010 average.

From the end of June, the weather in Greenland changed markedly, as warm, sunny conditions took hold. 

During the first 20 days of July, melting was well above the 1981-2020 average. At its peak on 22 July, melting covered 74.9% of the ice sheet. 

(For comparison, July 2025 holds the record for melt area in Greenland, when 81.2% of the continent was melting simultaneously.) 

However, the early season fresh snow in June helped to form a cold and bright buffer, reducing the subsequent summer ice melt and runoff.

The map below shows how roughly three-quarters of Greenland’s ice sheet – shaded in red – was experiencing surface melt on 22 July. The chart shows how surface melt on that date (blue line) outpaced the 1981-2010 average (grey line and shading).

Map showing areas of Greenland undergoing surface melt (shaded red) on 22 July 2026. Chart shows percentage of ice sheet area which saw surface melt on 22 July 2026 (blue) compared to the 1981-2010 average (grey). Credit: Polar Portal.
Snowfall

As mentioned above, snowfall is the only way for the ice sheet to gain mass.

Because the dry 2025-26 winter did not provide the usual thick layer of snow to act as a protective buffer, the ice sheet started the summer already “behind” on ice levels.

However, even though there was above-average melting during summer, the surface mass balance remained almost constant from mid-June to mid-July. It was only slightly smaller by the end of August.

This was because much of the melt trickled into the snow below and refroze.

South-east Greenland is typically the wettest part of the ice sheet and this year was no exception. Unlike other years though, there was also significant snowfall on the west coast north of Nuussuaq. 

Summer snowfall helps to preserve the ice sheet, as fresh snow is brighter than old glacier ice and reflects more of the sun’s energy back to space. 

Increases in both snowfall and melt are in line with what scientists expect in a warming climate. Warmer air holds more water vapour, bringing more snow and rain, while higher temperatures drive more ice melt. 

Iceberg calving 

The most dramatic ice loss of 2025-26 happened at Greenland’s edges.

On 4 August, a vast tabular iceberg – an “ice island” with nearly vertical sides and a flat top – broke away from the floating ice tongue of Petermann glacier in north-west Greenland. 

Sentinel-1 image sequence of the Petermann ice island, before and after the 4 August 2026 calving. Credit: ESA/Copernicus.

More than 76 square kilometres (km2) in size, with an estimated thickness of 150 metres, the amount of water frozen in this iceberg would be enough to submerge the entirety of greater London to a depth of 6.65m. 

This was the largest loss of floating ice at the Petermann glacier since 2012 and the biggest calving event by any Arctic glacier since 2020. 

Similar breakoffs occurred in 2010 and 2012 and the current one had long been expected, given growing cracks in the glacier for several years. 

Scientists anticipate more calving events at the Petermann glacier. Two further ice islands, with estimated areas of roughly 94 and 84km2, are expected to detach as existing rifts continue to cut across the ice tongue. 

Together, the three would remove about 22% of Petermann’s remaining floating ice.

Because the ice tongue is already floating, this calving does not directly raise sea levels. But Petermann is one of Greenland’s largest marine-terminating glaciers, acting as a brake on ice flowing from the interior to the ocean. 

The recent breakaway is a reminder that calving and glacier retreat continue apace in Greenland, even in years when the surface mass budget is unremarkable. (To date, calving and glacier retreat have driven roughly two-thirds of total ice loss on the continent.)

The impact of global weather

Greenland’s relatively high surface mass balance in 2025-26 can be explained, in part, by the hot summer in Europe.

Multiple heatwaves were observed across southern Europe and North America over the summer of 2026, with temperature records recorded in multiple countries.  

When high-pressure systems that cause heatwaves become blocked in place, the jet stream – fast-moving winds that blow from west to east high in the atmosphere – is shaped like the Greek capital letter Omega (Ω). This is the pattern that dominated this year.

Warm conditions settle near the cores of high-pressure systems, while cool, wet conditions can be found at the “feet” of the Omega. 

This year, Greenland was on the right side of the Omega-shaped jet stream and thus experienced cooler conditions. 

If the Omega pattern had been displaced by half a wavelength, Greenland would have likely seen warm temperatures that brought record levels of melt – and a lower surface mass balance. 

Three decades of decline

So, why does the milestone of 30 years of ice mass decline in Greenland matter? 

A single year, however dramatic, can be strongly influenced by the weather. An extreme melt summer – or an unusually cold, snowy one – says little on its own about the region’s  long-term trajectory. 

This is why climate scientists typically look at 30-year periods when distinguishing a change in climate from natural variability.

Thirty years of data shows that the annual ice budget in Greenland has ended in the red, without exception, since 1996. This applies to warm years, cold years and snowy years.

The largest single-year loss remains 2012, when around 458bn tonnes of ice disappeared from the ice sheet. 

The past two years have been moderate by comparison – yet even these “quiet” years produced losses of more than 100bn tonnes each.

Thirty years of data on ice mass decline means that scientists now can confidently say that Greenland mass loss is due to climate change. 

Greenland in the global sea level picture

A new assessment highlights how Greenland’s losses are the largest single contribution from the ice sheets to global sea level rise. 

The Ice Sheet Mass Balance Intercomparison Exercise (IMBIE) combines 42 independent estimates of ice sheet mass change from satellite gravimetry, altimetry and the input-output method – including the Danish Meteorological Institute’s surface mass balance estimates detailed in this article – all into a single record.

The latest assessment, published in the journal Scientific Data in September, shows that Greenland and Antarctica together lost 11,309bn tonnes of ice between 1979 and 2023, raising global sea levels by around 32mm. 

The two ice sheets now account for almost one-quarter of observed global sea level rise.

Greenland dominates that record. It lost 6,215bn tonnes of ice between 1972 and 2023 – an average of 119bn tonnes per year. Two-thirds of the loss was driven by increased ice discharge and one-third by reduced surface mass balance due to increased melt. 

Annual losses from ice melt accelerated from around 60bn tonnes in the 1980s to 264bn tonnes in the 2010s.

In contrast, Antarctica lost 4,780bn tonnes over the same period, raising sea levels by around 13mm. 

All of Antarctica’s losses are due to discharge from the ice sheet, concentrated in west Antarctica. East Antarctica, on the other hand, has so far remained relatively stable, with record snowfall in recent years even producing a small mass gain. 

This division in sea-level contribution is expected to change. Greenland’s losses to date amount to only a few centimetres of sea level rise, but Antarctica’s ice sheet is roughly 10 times larger and therefore could lead to significantly more sea level rise in the future. 

The IMBIE assessment notes that Antarctica’s losses are dominated by marine-terminating glaciers whose stability is hardest to predict.

The processes now increasingly observed in Antarctica – increasing surface melt, shrinking sea ice, loss of ice shelves and seasonal patterns of retreat of outlet glaciers – have been seen in Greenland for decades. We call this the “Greenlandification” of Antarctica. 

Greenland, in other words, is not just the largest present-day source of ice-driven sea level rise – it is also a preview of what may come from the far larger ice sheet at the other end of the planet.

Related Q&A: How Trump is threatening climate science in Earth’s polar regions 20.02.2026 Antarctica Guest post: The challenges in projecting future global sea levels 17.02.2026 Greenland State of the climate: 2025 in top-three hottest years on record as ocean heat surges 14.01.2026 Climate modelling Guest post: How the Greenland ice sheet fared in 2025 15.12.2025 Greenland

The post Guest post: Greenland ice sheet shrinks for 30th year in a row appeared first on Carbon Brief.

Categories: I. Climate Science

Chart: A record two-thirds of car sales in China are now electric vehicles

Fri, 10/23/2026 - 08:55
Print article Share Article Contents

For the first time, new-energy vehicles (NEVs) made up 67% of new car sales in China in September, according to the latest data from the China Passenger Car Association (CPCA).

The figures show that NEVs, which include battery electric cars and plug-in hybrids, captured a record share of sales, even though fewer were sold than last year.

This is because sales of combustion-engine cars are falling steeply, a trend that the CPCA has attributed to rising petrol prices following the start of the Hormuz crisis.

Indeed, sales of fuel-only cars were down 31% year-on-year in the first nine months of 2026, with plug-in hybrid sales down 26% – whereas battery EV sales were only down 4%.

One analyst tells Carbon Brief that China’s rapidly growing EV fleet has offset demand for 143m barrels of oil since the start of the Hormuz crisis.

Another says that the crisis is “reinforcing China’s electrification of road transport”.

Meanwhile, amid slowing domestic sales, Chinese carmakers’ exports of NEVs more than doubled [in 2026 to date / in September] from a year earlier.

NEV record

A record two out of every three cars sold in September in China were NEVs, according to CPCA data, as shown in the figure below.

(NEVs predominantly include battery electric vehicles – BEVs – and plug-in hybrids  – PHEVs. BEVs only run on electricity, whereas PHEVs also have a combustion engine.)

This marks a rapid rise for NEVs from just 6% of new car sales in 2020 up to 58% in 2026 to date – and a record 67% in the month of September.

This year, NEVs’ share of sales has also bounced back from a low of 39% in January, after the full purchase tax exemption that had been in place for more than a decade came to an end in December 2025.

The gain in NEV share in 2026 has come amid a slowdown across the wider domestic passenger-car market, which has hit sales of combustion-engine vehicles particularly hard. 

This September, passenger-car retail sales across all categories declined, with overall sales volumes falling 28% year-on-year.

The contraction was led by combustion-engine vehicles, which saw sales fall 43% year-on-year in September and 31% in 2026 to date.

However, sales of PHEVs also fell by 36% year-on-year in September and by 26% in 2026 to date. By comparison, BEVs performed better, although sales still declined 8.5% from a year earlier in September and 4% in 2026 to date.

As a result, NEV sales overall fell by 17% in September and 13% in 2026 to date.

While the market changes in 2026 to date were particularly dramatic, they reflect a broad continuation of trends seen over the past decade.

Indeed, sales of combustion-engine cars in China peaked in 2017 and they have been declining ever since, shown in the figure below.

Impact of Hormuz

Part of the reason for the advance of NEVs in 2026, said Chinese technology news outlet OFweek, was petrol cars “giving up market share” as they were outperformed by electric alternatives.

Cui Dongshu, secretary-general of the CPCA, cited higher petrol prices as an explanation for September’s weak combustion-car sales, in comments to financial news outlet National Business Daily. 

Imports of oil into China in January-September have fallen 15% compared to 2025, equivalent to 470m barrels, as ongoing conflict tightened supply and drove up prices. 

China’s passenger-car NEV fleet has displaced 143m barrels of oil so far during the strait of Hormuz crisis, Euan Graham, senior energy analyst at thinktank Ember, tells Carbon Brief.

EVs “accounted for 12% of the decrease in China’s oil use post-Hormuz”, according to an estimate provided to thinktank the Brookings Institute by an unnamed analyst.

#gform_wrapper_6[data-form-index="0"].gform-theme,[data-parent-form="6_0"]{--gf-color-primary: #204ce5;--gf-color-primary-rgb: 32, 76, 229;--gf-color-primary-contrast: #fff;--gf-color-primary-contrast-rgb: 255, 255, 255;--gf-color-primary-darker: #001AB3;--gf-color-primary-lighter: #527EFF;--gf-color-secondary: #fff;--gf-color-secondary-rgb: 255, 255, 255;--gf-color-secondary-contrast: #112337;--gf-color-secondary-contrast-rgb: 17, 35, 55;--gf-color-secondary-darker: #F5F5F5;--gf-color-secondary-lighter: #FFFFFF;--gf-color-out-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-out-ctrl-light-rgb: 17, 35, 55;--gf-color-out-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-out-ctrl-light-lighter: #F5F5F5;--gf-color-out-ctrl-dark: #585e6a;--gf-color-out-ctrl-dark-rgb: 88, 94, 106;--gf-color-out-ctrl-dark-darker: #112337;--gf-color-out-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-color-in-ctrl: #fff;--gf-color-in-ctrl-rgb: 255, 255, 255;--gf-color-in-ctrl-contrast: #112337;--gf-color-in-ctrl-contrast-rgb: 17, 35, 55;--gf-color-in-ctrl-darker: #F5F5F5;--gf-color-in-ctrl-lighter: #FFFFFF;--gf-color-in-ctrl-primary: #204ce5;--gf-color-in-ctrl-primary-rgb: 32, 76, 229;--gf-color-in-ctrl-primary-contrast: #fff;--gf-color-in-ctrl-primary-contrast-rgb: 255, 255, 255;--gf-color-in-ctrl-primary-darker: #001AB3;--gf-color-in-ctrl-primary-lighter: #527EFF;--gf-color-in-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-in-ctrl-light-rgb: 17, 35, 55;--gf-color-in-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-in-ctrl-light-lighter: #F5F5F5;--gf-color-in-ctrl-dark: #585e6a;--gf-color-in-ctrl-dark-rgb: 88, 94, 106;--gf-color-in-ctrl-dark-darker: #112337;--gf-color-in-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-radius: 3px;--gf-font-size-secondary: 14px;--gf-font-size-tertiary: 13px;--gf-icon-ctrl-number: url("data:image/svg+xml,%3Csvg width='8' height='14' viewBox='0 0 8 14' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M4 0C4.26522 5.96046e-08 4.51957 0.105357 4.70711 0.292893L7.70711 3.29289C8.09763 3.68342 8.09763 4.31658 7.70711 4.70711C7.31658 5.09763 6.68342 5.09763 6.29289 4.70711L4 2.41421L1.70711 4.70711C1.31658 5.09763 0.683417 5.09763 0.292893 4.70711C-0.0976311 4.31658 -0.097631 3.68342 0.292893 3.29289L3.29289 0.292893C3.48043 0.105357 3.73478 0 4 0ZM0.292893 9.29289C0.683417 8.90237 1.31658 8.90237 1.70711 9.29289L4 11.5858L6.29289 9.29289C6.68342 8.90237 7.31658 8.90237 7.70711 9.29289C8.09763 9.68342 8.09763 10.3166 7.70711 10.7071L4.70711 13.7071C4.31658 14.0976 3.68342 14.0976 3.29289 13.7071L0.292893 10.7071C-0.0976311 10.3166 -0.0976311 9.68342 0.292893 9.29289Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-select: url("data:image/svg+xml,%3Csvg width='10' height='6' viewBox='0 0 10 6' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M0.292893 0.292893C0.683417 -0.097631 1.31658 -0.097631 1.70711 0.292893L5 3.58579L8.29289 0.292893C8.68342 -0.0976311 9.31658 -0.0976311 9.70711 0.292893C10.0976 0.683417 10.0976 1.31658 9.70711 1.70711L5.70711 5.70711C5.31658 6.09763 4.68342 6.09763 4.29289 5.70711L0.292893 1.70711C-0.0976311 1.31658 -0.0976311 0.683418 0.292893 0.292893Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-search: url("data:image/svg+xml,%3Csvg width='640' height='640' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath d='M256 128c-70.692 0-128 57.308-128 128 0 70.691 57.308 128 128 128 70.691 0 128-57.309 128-128 0-70.692-57.309-128-128-128zM64 256c0-106.039 85.961-192 192-192s192 85.961 192 192c0 41.466-13.146 79.863-35.498 111.248l154.125 154.125c12.496 12.496 12.496 32.758 0 45.254s-32.758 12.496-45.254 0L367.248 412.502C335.862 434.854 297.467 448 256 448c-106.039 0-192-85.962-192-192z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-label-space-y-secondary: var(--gf-label-space-y-md-secondary);--gf-ctrl-border-color: #686e77;--gf-ctrl-size: var(--gf-ctrl-size-md);--gf-ctrl-label-color-primary: #112337;--gf-ctrl-label-color-secondary: #112337;--gf-ctrl-choice-size: var(--gf-ctrl-choice-size-md);--gf-ctrl-checkbox-check-size: var(--gf-ctrl-checkbox-check-size-md);--gf-ctrl-radio-check-size: var(--gf-ctrl-radio-check-size-md);--gf-ctrl-btn-font-size: var(--gf-ctrl-btn-font-size-md);--gf-ctrl-btn-padding-x: var(--gf-ctrl-btn-padding-x-md);--gf-ctrl-btn-size: var(--gf-ctrl-btn-size-md);--gf-ctrl-btn-border-color-secondary: #686e77;--gf-ctrl-file-btn-bg-color-hover: #EBEBEB;--gf-field-img-choice-size: var(--gf-field-img-choice-size-md);--gf-field-img-choice-card-space: var(--gf-field-img-choice-card-space-md);--gf-field-img-choice-check-ind-size: var(--gf-field-img-choice-check-ind-size-md);--gf-field-img-choice-check-ind-icon-size: var(--gf-field-img-choice-check-ind-icon-size-md);--gf-field-pg-steps-number-color: rgba(17, 35, 55, 0.8);}

CompanyThis field is for validation purposes and should be left unchanged.

Free fortnightly newsletter

.preheader p{ margin-top: 0; font-family: 'PT Sans', sans-serif; font-weight: var(--type--3--font-weight--bold); color: var(--button--color); font-size: var(--button--font-size, inherit); } .newsletter-inline{ display: flex; border: solid 1px #333333; padding: 1em; background: #ffffff; } .inline-email{ display:inline-block; margin-right:1em; margin-top:0 !important; margin-bottom:0.5em; } #field_submit{ display:inline-block; margin-top:0 !important; } .gform_wrapper .gfield+.gfield{ margin-top:0 } Email gform.initializeOnLoaded( function() {gformInitSpinner( 6, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery('#gform_ajax_frame_6').on('load',function(){var contents = jQuery(this).contents().find('*').html();var is_postback = contents.indexOf('GF_AJAX_POSTBACK') >= 0;if(!is_postback){return;}var form_content = jQuery(this).contents().find('#gform_wrapper_6');var is_confirmation = jQuery(this).contents().find('#gform_confirmation_wrapper_6').length > 0;var is_redirect = contents.indexOf('gformRedirect(){') >= 0;var is_form = form_content.length > 0 && ! is_redirect && ! is_confirmation;var mt = parseInt(jQuery('html').css('margin-top'), 10) + parseInt(jQuery('body').css('margin-top'), 10) + 100;if(is_form){jQuery('#gform_wrapper_6').html(form_content.html());if(form_content.hasClass('gform_validation_error')){jQuery('#gform_wrapper_6').addClass('gform_validation_error');} else {jQuery('#gform_wrapper_6').removeClass('gform_validation_error');}setTimeout( function() { /* delay the scroll by 50 milliseconds to fix a bug in chrome */ jQuery(document).scrollTop(jQuery('#gform_wrapper_6').offset().top - mt); }, 50 );if(window['gformInitDatepicker']) {gformInitDatepicker();}if(window['gformInitPriceFields']) {gformInitPriceFields();}var current_page = jQuery('#gform_source_page_number_6').val();gformInitSpinner( 6, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery(document).trigger('gform_page_loaded', [6, current_page]);window['gf_submitting_6'] = false;}else if(!is_redirect){var confirmation_content = jQuery(this).contents().find('.GF_AJAX_POSTBACK').html();if(!confirmation_content){confirmation_content = contents;}jQuery('#gform_wrapper_6').replaceWith(confirmation_content);jQuery(document).scrollTop(jQuery('#gf_6').offset().top - mt);jQuery(document).trigger('gform_confirmation_loaded', [6]);window['gf_submitting_6'] = false;wp.a11y.speak(jQuery('#gform_confirmation_message_6').text());}else{jQuery('#gform_6').append(contents);if(window['gformRedirect']) {gformRedirect();}}jQuery(document).trigger("gform_pre_post_render", [{ formId: "6", currentPage: "current_page", abort: function() { this.preventDefault(); } }]); if (event && event.defaultPrevented) { return; } const gformWrapperDiv = document.getElementById( "gform_wrapper_6" ); if ( gformWrapperDiv ) { const visibilitySpan = document.createElement( "span" ); visibilitySpan.id = "gform_visibility_test_6"; gformWrapperDiv.insertAdjacentElement( "afterend", visibilitySpan ); } const visibilityTestDiv = document.getElementById( "gform_visibility_test_6" ); let postRenderFired = false; function triggerPostRender() { if ( postRenderFired ) { return; } postRenderFired = true; gform.core.triggerPostRenderEvents( 6, current_page ); if ( visibilityTestDiv ) { visibilityTestDiv.parentNode.removeChild( visibilityTestDiv ); } } function debounce( func, wait, immediate ) { var timeout; return function() { var context = this, args = arguments; var later = function() { timeout = null; if ( !immediate ) func.apply( context, args ); }; var callNow = immediate && !timeout; clearTimeout( timeout ); timeout = setTimeout( later, wait ); if ( callNow ) func.apply( context, args ); }; } const debouncedTriggerPostRender = debounce( function() { triggerPostRender(); }, 200 ); if ( visibilityTestDiv && visibilityTestDiv.offsetParent === null ) { const observer = new MutationObserver( ( mutations ) => { mutations.forEach( ( mutation ) => { if ( mutation.type === 'attributes' && visibilityTestDiv.offsetParent !== null ) { debouncedTriggerPostRender(); observer.disconnect(); } }); }); observer.observe( document.body, { attributes: true, childList: false, subtree: true, attributeFilter: [ 'style', 'class' ], }); } else { triggerPostRender(); } } );} );

Erica Downs, a senior research scholar at Columbia University’s Center on Global Energy Policy, tells Carbon Brief that the strait of Hormuz crisis and higher oil prices are “reinforcing China’s electrification of road transport”. 

She says the “world’s largest oil supply disruption” vindicates China’s early bet on EVs, which was partly driven by concerns about oil-supply security.

NEV demand down 

Nevertheless, NEV sales have fallen year-on-year, albeit at a slower rate than overall car sales. The growth in NEV market share has also been decelerating.

Zhang Yichao, a partner at consulting firm AlixPartners, tells Carbon Brief that NEVs benefit from government trade-in policies and a “strong pipeline” of new models, particularly those involving innovation. 

He says these factors help explain why NEVs sales have fallen at a slower rate to their combustion-car counterparts.

Cui told the state-run China National Radio that the fall for NEVs and for cars overall in September 2026, is also due to a high baseline of sales from the same month in 2025, as well as changes to trade-in policies and disruptions from the holiday period. 

Other factors, says consulting firm Wood Mackenzie, are the weakened demand due to “demand pull-forward” in 2025 caused by the end of subsidies, as well as “tighter financing conditions”.

Zhang tells Carbon Brief he does not “see much more room ahead” for further growth in NEVs’ market share.

He adds that, in his view, their share in China is unlikely to exceed 80%, as some premium-market segments and specific use cases will continue to require combustion-engine vehicles.

Exports surge

Meanwhile, while domestic NEV retail sales fell 17% in September 2026, exports more than doubled, increasing 164% year-on-year, according to the CPCA.

This extends the decline in domestic sales and very strong growth in exports seen across 2026 to date, as shown in the figure below.

In September, 1,084,000 NEVs were sold domestically in China, while 558,000 NEVs were exported.

Exports are becoming increasingly important to Chinese carmakers, in the face of weaker domestic demand and cut-throat industrial competition.

The CPCA says exports have become “a core priority” for some carmakers.

However, the surge in exports from China has created concern in the EU, which has placed tariffs on imports of EVs from China and recently agreed plans to cut imports of hybrids.

Countries including Mexico and Turkey have also placed tariffs on Chinese EV imports. 

At the same time, the Chinese government is putting policies in place that might be expected to disincentivise exports, such as cutting VAT export rebates.

Analysts differ on what impact this policy shift will have, with some saying it will be limited, while others believe it could cause the price of exported cars to rise. 

Related Experts: What China wants from the COP31 climate talks 29.10.2026 China policy Analysis: China’s clean-tech exports in 2026 to avoid as much CO2 as Brazil emits in a year 29.10.2026 China policy Interactive: What China’s dozens of ‘five-year plans’ say about climate and energy 24.09.2026 China policy Q&A: What does China’s 15th five-year plan for coal mean for climate action? 14.08.2026 China policy

The post Chart: A record two-thirds of car sales in China are now electric vehicles appeared first on Carbon Brief.

Categories: I. Climate Science

Experts: How are climate scientists using artificial intelligence?

Thu, 10/22/2026 - 03:45
Print article Share

Artificial intelligence (AI) is developing at a rapid pace and is increasingly being used by scientists to better understand climate change.

AI is perhaps best-known for its use in “large language models”, such as ChatGPT, Claude and Gemini.

However, the field is vast. 

From weather forecasting to analysing satellite images, climate scientists are using various types of AI to develop faster and more efficient ways to study the warming planet. 

However, the technology comes with no shortage of risks and challenges, including overreliance on unverified results, intensive energy use and exacerbating existing inequalities in climate science.

Carbon Brief spoke to 10 experts working across a range of climate-related fields – from developing climate forecasts to analysing changes in the energy grid – about how AI is being used.

Their answers are set out below, grouped by theme.

Article Contents What are the benefits and challenges of using AI in weather and climate forecasting?

Dr Doug McNeall, science lead in data-driven climate modelling at the UK Met Office:

One of the most significant recent developments has been the success of AI weather-prediction models, which learn from large volumes of weather data to predict the next few hours ahead and then repeatedly feed those predictions back into themselves to generate forecasts up to two weeks into the future.

These models have demonstrated that AI can provide accurate predictions while requiring far less computing power than existing numerical models [which rely on mathematical equations to simulate the climate].

One of the main challenges is that AI weather-prediction models have been developed for relatively short timescales, where large amounts of observational data [for example, temperature, rainfall or sea ice extent data] is available and model performance can be evaluated quickly and rigorously.

Climate prediction is fundamentally different. To provide useful climate information, we need to make projections for future climates that lie outside the observational record.

One promising route forward is, therefore, to use AI to emulate or accelerate these existing models, rather than relying solely on purely data-driven approaches. This blended approach could make climate risk information far more accessible and affordable to produce, while retaining the physical understanding [based on mathematical and scientific principles] that underpins confidence in climate projections.

Prof William Collins, professor of AI for climate science at the University of East Anglia:

We’re seeing a really exciting renaissance in both weather and climate science…Machine learning [a subset of AI that uses algorithms to let computers learn from data and make predictions without being explicitly programmed] has had an enormous impact on numerical weather prediction.

So, for example, the European Center for Mid-Range Weather Forecasts is now deploying an AI forecast system as part of their official communication of weather predictions to the European community, alongside the traditional forecast system. AI systems have become good enough that they’re absolutely competitive with the world’s leading numerical weather prediction…So that’s been quite remarkable.

Much of machine learning and AI involves learning from data and climate represents a future for which we do not have data. We’re literally trying to predict the future. We don’t have observations of it. So, where does the data come from to inform the models? That is, at the moment, the central challenge that the climate community is attempting to overcome.

Dr Amirpasha Mozaffari, researcher in Earth artificial intelligence at the Barcelona Supercomputing Centre:

AI forecast models do well on typical weather, but underperform physics-based models on record-breaking and unprecedented extremes – precisely the events that matter most for adaptation [to climate change].

Solving that – whether through hybrid models [which combine AI with traditional models], physical constraints [in which rules are implemented based on the laws of physics to help the models make realistic forecasts] or new ways of evaluating models on what they have not seen – is the central open problem for the field.

What are some other current uses of AI in climate science research?

Dr Stefano Materia, researcher in climate variability and change at the Barcelona Supercomputing Centre:

Besides the climate modelling and prediction framework, the proliferation of Earth-observation satellites has created a massive influx of geospatial data (for example, images showing deforestation or sea ice melt). AI is used to translate raw pixel data into actionable climate intelligence.

Methane tracking and ecosystem monitoring are two examples of this kind of application. Methane is a potent greenhouse gas, but leaks are largely invisible and hard to track. AI algorithms now automatically scan hyperspectral imagery [which uses data from across the electromagnetic spectrum] from satellites to pinpoint specific industrial methane leaks globally, alerting operators in near real-time.

At the same time, machine-learning models process thousands of satellite images daily to detect illegal logging, track land-use changes and monitor glacier retreat long before humans could manually review the footage.

Satellite image showing deforestation in Para, Brazil. Credit: Universal Images Group North America LLC / Alamy Stock Photo

Dr Clare Eayrs, polar research scientist at the Environmental Fluid Dynamics Lab at New York University:

In sea ice research, machine learning is now used across the sea ice modelling workflow…The most mature application is extracting more information from observations, for example, mapping fine-scale sea ice deformation from ship radar and classifying snow layers in micropenetrometer profiles [which measure the hardness or softness of different layers of snow].

Forecasting is the most visible application in this field. Ice Net [a sea ice forecasting system] was an early demonstration that a deep-learning system [which uses models inspired by the layers of a human brain to process information] could outperform a leading physical model in seasonal Arctic forecasts. Machine-learning systems are now used to predict [climate] variables such as sea ice concentration, thickness and drift.

Prof Cristiana Stan, vice-chair of the World Climate Research Programme (WCRP):

AI now appears across all major parts of the WCRP. [For example,]…the ”weather prediction model intercomparison project” [an initiative to create a centralised database of physics- based, machine-learning and hybrid model forecasts] includes physics-based numerical weather prediction models, hybrid [models] and AI weather prediction models. Because it is focused on operational models, the project offers a direct comparison between state-of-the art systems. [It] is intended to accelerate the deployment of the next generation of operational models and contributes to development of metrics needed to assess relative strengths and weaknesses of models.

Downscaling is another area that is harnessing AI power to refine the coarse-resolution model output into high-resolution.

Dr William Collins:

The Intergovernmental Panel on Climate Change (IPCC) is performing its seventh assessment report (AR7)…This time, for the very first time, there’s a model intercomparison project just for models using AI [for weather and climate forecasting]. So that is a real milestone…It’s a very exciting proof of principle or proof of concept and, I think, the harbinger of many more developments of this kind to come. 

[For more on how model intercomparison will inform IPCC AR7, see Carbon Brief’s recent guest post.]

Dr Amirpasha Mozaffari:

A few years ago, it took two or three years for a new architecture to make its way from the machine-learning community into weather and climate work. Today it is a matter of months: diffusion models, foundation models and generative ensembles have all arrived in our field almost as soon as they appeared elsewhere. That speed opens up things that were previously out of reach.

And the cost of running a global forecast has dropped by orders of magnitude…which, in principle, changes who can produce weather and climate information at all. 

How is AI being used to advance climate policy and action?

Prof Priya Donti, assistant professor at the Massachusetts Institute of Technology and co-founder of Climate Change AI:

As we introduce more solar, wind and other renewable energy into the power grid, we need to rethink how we actually manage the various sources on the power grid… 

AI is used in a few different ways here. One of them is just trying to get better predictions of solar and wind power production in order to reduce that uncertainty and allow us to plan with a little more foresight.

Wind turbines and solar panels in California. Credit: Tom Grundy / Alamy Stock Photo

Another big one is helping to actually optimise and control the various components we have on power grids – the power generators, the batteries, flexible loads like industrial loads or the batteries in your electric vehicles or heating and cooling systems.

Dr Carlos Rodriguez-Pardo, postdoctoral researcher at Politecnico di Milano, affiliated scientist at the European Institute on Economics and the Environment and at the Euro-Mediterranean Center on Climate Change:

AI is already useful at every stage of the climate policy cycle. 

The first stage is measurement: understanding what is happening, what has been tried and who is exposed. Machine learning helps us process satellite imagery to estimate emissions, improve air quality monitoring, harmonise fragmented policy databases and map a scientific literature that has grown far beyond what any human team could read.

The second stage is modelling futures. The models we use to explore climate and economic scenarios are computationally very expensive, so AI-based emulators [simple models that can be used to represent more complex mathematical models] that reproduce their behaviour at a small fraction of the cost are transforming what we can do. Instead of a handful of scenarios, we can explore thousands and quantify uncertainties that were previously out of reach.

Reinforcement learning [a type of AI in which a computer programme learns to make decisions through a process of trial and error] is also allowing us to study how multiple actors, such as countries or electricity producers, compete and cooperate under different policy designs. This can reveal strategies that classical analysis misses. 

The third stage is evaluation, where causal machine learning is extending the toolkit we use to answer whether a policy actually worked. That is a different – and much harder – question than predicting what will happen next.

How could AI be used in climate research over the coming years?

Prof Cristiana Stan:

The WCRP has established a task team to address this very question. The focus areas considered by the task team include modelling and projections, data and observations [and AI] impacts and services, engagement with the private sector supporting development and integration of AI and machine learning into climate research and applications and the scientific rigour and trustworthiness of AI and machine learning methods.

Dr William Collins:

I think the IPCC’s eighth assessment [likely due in the late 2020s] may itself largely be written using large language models. Basically, the job of the people writing the IPCC report is to synthesise…One could certainly argue whether or not that needs to be done by people. AI should be telling us the known unknowns, not the known knowns… The thing that’s value-added [by AI] as part of that assessment process is to tell us what isn’t certain, what isn’t known well. Because then that tells you where the science should be heading.

[In September 2026, the IPCC established an “informal group” on AI to “provide timely guidance on the responsible use of AI and emerging digital technologies in IPCC processes”.]

AI will also be an integral part of the eighth Coupled Model Intercomparison Project, with machine-learning codes running alongside traditional climate models.

Dr Stefano Materia:

The inclusion of AI in climate and land-vegetation modelling represents a paradigm shift. Rather than replacing physical models entirely, hybrid modelling merges the mathematical rigour and physical consistency of traditional process-based Earth system models [advanced computer models that combine physical, chemical and biological processes to simulate the climate] with the pattern-recognition power and speed of machine learning.

This new approach will be able to overcome computational bottlenecks, replace unresolved subgrid processes [which are too small to be captured by the grid cells of a computer model], improve the terrestrial biosphere – finally narrowing down the uncertainties on the terrestrial carbon sink and accurately pinpointing emergent constraints and non-linear tipping points.

Dr Clare Eayrs:

Much of what we know about Antarctic sea ice comes from the satellite era, which begins in 1979. That is a short record for a system with strong decadal variability. A great deal of information about earlier conditions is contained in ship logs, whaling records, expedition photographs and early aerial and pre-1979 satellite imagery. Machine learning is well suited to extracting it. Lengthening the observational record would improve almost everything else we do, including training and testing AI systems.

“The Bear” whaling ship getting free from the ice pack. Credit: New Bedford Whaling Museum Are there any misconceptions about the use of AI in climate science?

Dr David Rolnick:

AI is frequently thought of as large generative models, like [those] developed by large tech companies. But that is not mostly what is useful in AI for tackling climate change.

People talk a lot about energy usage of AI and negative impacts on society, including on the climate, as a result of that energy usage and associated negative impacts like water usage. These are very real worries. However, it is worth noting that the massive energy and other requirements of AI systems are referring to those large-scale generative and agentic systems. They’re not mostly about the kinds of AI systems that are actually used to fight climate change.

Prof Priya Donti:

I think the general public thinks of AI as this one specific thing…often large language models, chatbots, etc…People who are making allocation decisions for who funds scientists might have a mismatched understanding of what needs to happen, which means the right science doesn’t necessarily get funded.

Could AI improve or exacerbate inequalities in climate research?

Dr Amirpasha Mozzafari: 

Frontier models [which are at the cutting edge of AI research] are built almost exclusively in the global north, trained on data that is densest where observations are richest and validated against benchmarks that reflect the same regions. 

Cheap inference [in which a trained AI model applies its learned patterns to new, unseen data] does not fix that on its own: being able to run someone else’s model is not the same as having a say in what it is trained on, how it is evaluated, or whether it serves your decisions. 

Without deliberate effort to share data infrastructure, evaluation metrics that centre on people’s wellbeing and genuine co-production of knowledge, AI risks automating and amplifying the existing divide in the global climate information system rather than closing it.

#gform_wrapper_5[data-form-index="0"].gform-theme,[data-parent-form="5_0"]{--gf-color-primary: #204ce5;--gf-color-primary-rgb: 32, 76, 229;--gf-color-primary-contrast: #fff;--gf-color-primary-contrast-rgb: 255, 255, 255;--gf-color-primary-darker: #001AB3;--gf-color-primary-lighter: #527EFF;--gf-color-secondary: #fff;--gf-color-secondary-rgb: 255, 255, 255;--gf-color-secondary-contrast: #112337;--gf-color-secondary-contrast-rgb: 17, 35, 55;--gf-color-secondary-darker: #F5F5F5;--gf-color-secondary-lighter: #FFFFFF;--gf-color-out-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-out-ctrl-light-rgb: 17, 35, 55;--gf-color-out-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-out-ctrl-light-lighter: #F5F5F5;--gf-color-out-ctrl-dark: #585e6a;--gf-color-out-ctrl-dark-rgb: 88, 94, 106;--gf-color-out-ctrl-dark-darker: #112337;--gf-color-out-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-color-in-ctrl: #fff;--gf-color-in-ctrl-rgb: 255, 255, 255;--gf-color-in-ctrl-contrast: #112337;--gf-color-in-ctrl-contrast-rgb: 17, 35, 55;--gf-color-in-ctrl-darker: #F5F5F5;--gf-color-in-ctrl-lighter: #FFFFFF;--gf-color-in-ctrl-primary: #204ce5;--gf-color-in-ctrl-primary-rgb: 32, 76, 229;--gf-color-in-ctrl-primary-contrast: #fff;--gf-color-in-ctrl-primary-contrast-rgb: 255, 255, 255;--gf-color-in-ctrl-primary-darker: #001AB3;--gf-color-in-ctrl-primary-lighter: #527EFF;--gf-color-in-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-in-ctrl-light-rgb: 17, 35, 55;--gf-color-in-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-in-ctrl-light-lighter: #F5F5F5;--gf-color-in-ctrl-dark: #585e6a;--gf-color-in-ctrl-dark-rgb: 88, 94, 106;--gf-color-in-ctrl-dark-darker: #112337;--gf-color-in-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-radius: 3px;--gf-font-size-secondary: 14px;--gf-font-size-tertiary: 13px;--gf-icon-ctrl-number: url("data:image/svg+xml,%3Csvg width='8' height='14' viewBox='0 0 8 14' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M4 0C4.26522 5.96046e-08 4.51957 0.105357 4.70711 0.292893L7.70711 3.29289C8.09763 3.68342 8.09763 4.31658 7.70711 4.70711C7.31658 5.09763 6.68342 5.09763 6.29289 4.70711L4 2.41421L1.70711 4.70711C1.31658 5.09763 0.683417 5.09763 0.292893 4.70711C-0.0976311 4.31658 -0.097631 3.68342 0.292893 3.29289L3.29289 0.292893C3.48043 0.105357 3.73478 0 4 0ZM0.292893 9.29289C0.683417 8.90237 1.31658 8.90237 1.70711 9.29289L4 11.5858L6.29289 9.29289C6.68342 8.90237 7.31658 8.90237 7.70711 9.29289C8.09763 9.68342 8.09763 10.3166 7.70711 10.7071L4.70711 13.7071C4.31658 14.0976 3.68342 14.0976 3.29289 13.7071L0.292893 10.7071C-0.0976311 10.3166 -0.0976311 9.68342 0.292893 9.29289Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-select: url("data:image/svg+xml,%3Csvg width='10' height='6' viewBox='0 0 10 6' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M0.292893 0.292893C0.683417 -0.097631 1.31658 -0.097631 1.70711 0.292893L5 3.58579L8.29289 0.292893C8.68342 -0.0976311 9.31658 -0.0976311 9.70711 0.292893C10.0976 0.683417 10.0976 1.31658 9.70711 1.70711L5.70711 5.70711C5.31658 6.09763 4.68342 6.09763 4.29289 5.70711L0.292893 1.70711C-0.0976311 1.31658 -0.0976311 0.683418 0.292893 0.292893Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-search: url("data:image/svg+xml,%3Csvg width='640' height='640' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath d='M256 128c-70.692 0-128 57.308-128 128 0 70.691 57.308 128 128 128 70.691 0 128-57.309 128-128 0-70.692-57.309-128-128-128zM64 256c0-106.039 85.961-192 192-192s192 85.961 192 192c0 41.466-13.146 79.863-35.498 111.248l154.125 154.125c12.496 12.496 12.496 32.758 0 45.254s-32.758 12.496-45.254 0L367.248 412.502C335.862 434.854 297.467 448 256 448c-106.039 0-192-85.962-192-192z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-label-space-y-secondary: var(--gf-label-space-y-md-secondary);--gf-ctrl-border-color: #686e77;--gf-ctrl-size: var(--gf-ctrl-size-md);--gf-ctrl-label-color-primary: #112337;--gf-ctrl-label-color-secondary: #112337;--gf-ctrl-choice-size: var(--gf-ctrl-choice-size-md);--gf-ctrl-checkbox-check-size: var(--gf-ctrl-checkbox-check-size-md);--gf-ctrl-radio-check-size: var(--gf-ctrl-radio-check-size-md);--gf-ctrl-btn-font-size: var(--gf-ctrl-btn-font-size-md);--gf-ctrl-btn-padding-x: var(--gf-ctrl-btn-padding-x-md);--gf-ctrl-btn-size: var(--gf-ctrl-btn-size-md);--gf-ctrl-btn-border-color-secondary: #686e77;--gf-ctrl-file-btn-bg-color-hover: #EBEBEB;--gf-field-img-choice-size: var(--gf-field-img-choice-size-md);--gf-field-img-choice-card-space: var(--gf-field-img-choice-card-space-md);--gf-field-img-choice-check-ind-size: var(--gf-field-img-choice-check-ind-size-md);--gf-field-img-choice-check-ind-icon-size: var(--gf-field-img-choice-check-ind-icon-size-md);--gf-field-pg-steps-number-color: rgba(17, 35, 55, 0.8);}

LinkedInThis field is for validation purposes and should be left unchanged.

Free fortnightly newsletter

.preheader p{ margin-top: 0; font-family: 'PT Sans', sans-serif; font-weight: var(--type--3--font-weight--bold); color: var(--button--color); font-size: var(--button--font-size, inherit); } .newsletter-inline{ display: flex; border: solid 1px #333333; padding: 1em; background: #ffffff; } .inline-email{ display:inline-block; margin-right:1em; margin-top:0 !important; margin-bottom:0.5em; } #field_submit{ display:inline-block; margin-top:0 !important; } .gform_wrapper .gfield+.gfield{ margin-top:0 } Email gform.initializeOnLoaded( function() {gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery('#gform_ajax_frame_5').on('load',function(){var contents = jQuery(this).contents().find('*').html();var is_postback = contents.indexOf('GF_AJAX_POSTBACK') >= 0;if(!is_postback){return;}var form_content = jQuery(this).contents().find('#gform_wrapper_5');var is_confirmation = jQuery(this).contents().find('#gform_confirmation_wrapper_5').length > 0;var is_redirect = contents.indexOf('gformRedirect(){') >= 0;var is_form = form_content.length > 0 && ! is_redirect && ! is_confirmation;var mt = parseInt(jQuery('html').css('margin-top'), 10) + parseInt(jQuery('body').css('margin-top'), 10) + 100;if(is_form){jQuery('#gform_wrapper_5').html(form_content.html());if(form_content.hasClass('gform_validation_error')){jQuery('#gform_wrapper_5').addClass('gform_validation_error');} else {jQuery('#gform_wrapper_5').removeClass('gform_validation_error');}setTimeout( function() { /* delay the scroll by 50 milliseconds to fix a bug in chrome */ jQuery(document).scrollTop(jQuery('#gform_wrapper_5').offset().top - mt); }, 50 );if(window['gformInitDatepicker']) {gformInitDatepicker();}if(window['gformInitPriceFields']) {gformInitPriceFields();}var current_page = jQuery('#gform_source_page_number_5').val();gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery(document).trigger('gform_page_loaded', [5, current_page]);window['gf_submitting_5'] = false;}else if(!is_redirect){var confirmation_content = jQuery(this).contents().find('.GF_AJAX_POSTBACK').html();if(!confirmation_content){confirmation_content = contents;}jQuery('#gform_wrapper_5').replaceWith(confirmation_content);jQuery(document).scrollTop(jQuery('#gf_5').offset().top - mt);jQuery(document).trigger('gform_confirmation_loaded', [5]);window['gf_submitting_5'] = false;wp.a11y.speak(jQuery('#gform_confirmation_message_5').text());}else{jQuery('#gform_5').append(contents);if(window['gformRedirect']) {gformRedirect();}}jQuery(document).trigger("gform_pre_post_render", [{ formId: "5", currentPage: "current_page", abort: function() { this.preventDefault(); } }]); if (event && event.defaultPrevented) { return; } const gformWrapperDiv = document.getElementById( "gform_wrapper_5" ); if ( gformWrapperDiv ) { const visibilitySpan = document.createElement( "span" ); visibilitySpan.id = "gform_visibility_test_5"; gformWrapperDiv.insertAdjacentElement( "afterend", visibilitySpan ); } const visibilityTestDiv = document.getElementById( "gform_visibility_test_5" ); let postRenderFired = false; function triggerPostRender() { if ( postRenderFired ) { return; } postRenderFired = true; gform.core.triggerPostRenderEvents( 5, current_page ); if ( visibilityTestDiv ) { visibilityTestDiv.parentNode.removeChild( visibilityTestDiv ); } } function debounce( func, wait, immediate ) { var timeout; return function() { var context = this, args = arguments; var later = function() { timeout = null; if ( !immediate ) func.apply( context, args ); }; var callNow = immediate && !timeout; clearTimeout( timeout ); timeout = setTimeout( later, wait ); if ( callNow ) func.apply( context, args ); }; } const debouncedTriggerPostRender = debounce( function() { triggerPostRender(); }, 200 ); if ( visibilityTestDiv && visibilityTestDiv.offsetParent === null ) { const observer = new MutationObserver( ( mutations ) => { mutations.forEach( ( mutation ) => { if ( mutation.type === 'attributes' && visibilityTestDiv.offsetParent !== null ) { debouncedTriggerPostRender(); observer.disconnect(); } }); }); observer.observe( document.body, { attributes: true, childList: false, subtree: true, attributeFilter: [ 'style', 'class' ], }); } else { triggerPostRender(); } } );} );

Dr Carlos Rodriguez-Pardo:

If done well, [developing AI-based Earth system models] could dramatically lower barriers to entry. A small research group, or an agency in a data-poor region, could adapt a pre-trained model on modest hardware rather than building bespoke systems from scratch. That matters because [the] data is thinnest and the resolution gap is largest exactly where vulnerability to climate [change] is highest.

Dr David Rolnick, associate professor in the school of computer science at McGill University and cofounder of Climate Change AI:

Climate action should fundamentally be about democratisation of action and equitable response to climate change, both in terms of mitigation and in terms of adaptation. The consolidation of large-scale AI can often work counter to those goals by consolidating power in certain geographies and stakeholders, and by defining action based upon the goals of those stakeholders. 

When you build technology in specific institutions, they do not necessarily understand the needs of the majority of the world’s population or local stakeholders from other geographies and other entities. 

What are some other risks and challenges of AI in climate change research?

Dr Clare Eayrs:

The biggest challenge is trust…Good performance is necessary, but it is not enough. We also need independent evaluation, physical credibility, process-based validation, robustness and some degree of interpretability.

The risk I worry about most is that the apparent success of these systems could be read as a reason to invest less on observations. AI is not a substitute for observation, physical understanding or careful evaluation. Especially for Antarctic sea ice, where observations of sea ice and snow thickness remain sparse and understanding of snow-ice processes remains limited, we need more, targeted observations, not fewer. It also makes the case for documenting them – and their uncertainties – far better than we do now.

Dr William Collins:

The challenge is always making sure that you’re getting the right answer for the right reason and the risk is getting a plausible answer that is “truthy”, but not the truth…Ask a computer for an answer and it will give you one. And it’s very good. It will always obey. It will always give you an answer. The question is whether or not that answer is valid or not.

Dr David Rolnick:

The first and foremost risk is AI diverging attention from other forms of climate action. AI is not a silver bullet. It is not an answer in itself. And where it is useful, it is only one part of the solution.

Prof Detlef van Vuuren, a senior researcher at the PBL Netherlands Environmental Assessment Agency [in a 2026 Carbon Brief interview]:

There are now many more proposals being submitted than in the past, as people can use AI tools. So our science discipline has to start learning how to deal with this, both in terms of what we feel is allowed in creating [scientific] articles – what is allowed in reviewing [and] what is allowed in proposal writing. This is something that we have to learn.

Dr William Collins:

The partnership between academia, the data centres and private industry is going to be integral to our success going forward. And I think the challenge is going to be navigating that relationship in a way that respects open science, uses open source code, etc…

You do not want to be in a situation where we have something like a large language model where it’s not open source…[We need] fairness, accountability, transparency and explainability. 

Dr Carlos Rodriguez-Pardo:

High-resolution outputs can look exact when they are not; uncertainties compound in invisible ways and predictions that feed into policy need honest, well-calibrated uncertainty estimates, communicated in decision-relevant terms.

Like any other technology, AI should augment human judgement in climate decisions, not replace it. As researchers, it is our responsibility to ensure the field serves as a force for social progress.

Related Explainer: The CMIP7 emissions scenarios – and how they explore future climate change 01.09.2026 Climate modelling Guest post: How CMIP7 will shape the next wave of climate science 22.05.2026 Climate modelling Traditional models still ‘outperform AI’ for extreme weather forecasts 29.04.2026 Climate modelling Limiting global warming to 2C would not ‘rule out’ extreme impacts 25.03.2026 Climate modelling

The post Experts: How are climate scientists using artificial intelligence? appeared first on Carbon Brief.

Categories: I. Climate Science

Revealed: Expensive gas behind ‘nearly 90%’ of rise in UK electricity bills

Wed, 10/21/2026 - 04:37
Print article Share

A surge in wholesale costs driven by expensive gas is responsible for nearly 90% of the rise in UK household electricity bills since 2020, reveals new analysis.

The figures, shared with Carbon Brief by thinktank Nesta, show that wholesale gas costs are also behind the vast majority (77%) of the increase in household gas bills.

The analysis firmly contradicts the idea, which has become embedded in UK media and political narratives, that the UK’s net-zero target is to blame for high energy bills.

Andrew Sissons, director for sustainable future at Nesta, tells Carbon Brief that net-zero is, in fact, “helping to shelter Britain from the impact of expensive gas”. He adds: 

“Rising gas prices, not net-zero, have driven most of the recent surge in Britain’s energy bills. Britain is particularly exposed, with over 80% of households connected to gas mains, while our electricity system still relies too much on gas as the marginal unit.”

The analysis comes after the UK government gave its approval for the Jackdaw gas field to be developed in the North Sea, despite stating that it “would not…affect prices”.

Another gas crisis

In January 2026, Carbon Brief reported that two-thirds of the rise in electricity bills since before the 2022 global energy crisis was due to high gas prices.

Now, amid ongoing conflict in the Middle East, wholesale gas prices are at their highest levels since Russia’s 2022 invasion of Ukraine.

Indeed, as of 19 October, UK wholesale gas prices stood at 220p per therm (£75 per megawatt hour, MWh), above the average for 2022 of 210p/therm (£72/MWh).

The figure below shows that, after accounting for inflation, typical UK annual household electricity bills had risen £180 by the fourth quarter of 2026, relative to the same period in 2020.

Some 90% of this increase (£161) is due to wholesale costs, driven by the price of gas.

In contrast – and contrary to widespread perceptions – renewable “contracts for difference” (CfDs) are reducing electricity bills by £17 under the current price cap, according to the Nesta analysis.

In addition, the government has cut policy costs for electricity by moving 75% of legacy renewable support onto general taxation, ending the “ECO” efficiency scheme and temporarily removing VAT.

Network costs have added £72 to typical bills since 2020, according to Nesta. Most of this (£46) is for investment in the electricity transmission network, known as the “motorways” of the grid.

Only £21 of the rise in network costs relates to balancing the system, including “constraint payments” for turning off windfarms – and turning on gas plants – due to insufficient grid capacity.

Wind constraint payments regularly receive extensive news coverage, despite balancing costs – in total – making up just £35 (4%) of a typical £858 annual household electricity bill.

The Nesta figures show energy supplier operating costs have gone up by £46 since 2020. These costs are not broken down in the underlying Ofgem figures, but the increase is thought to be driven by the cost of rising household energy debt. Supplier profits are capped at 2.5% of bills.

Breaking the link

Notably, household electricity prices have increased far less rapidly since the Hormuz crisis than gas bills, with higher clean-energy generation helping break the link between the two.

Since 2020, household gas bills have increased by £368 in real terms, according to Nesta’s analysis. Some 77% of this (£283) is due to wholesale gas costs.

After accounting for inflation, total dual-fuel bills for a typical household have increased from £1,175 per year in October 2020 to £1,723 per year in October 2026, a rise of 47%.

The rise in wholesale costs driven by high gas prices accounts for £444 (81%) of this increase in real terms, as shown in the figure below.

The £127 real-terms rise in network costs includes £72 for electricity transmission, distribution and balancing, as well as £55 for investment in the gas grid.

(According to the Energy Networks Association, up to 80% of the investment in electricity grids is required to maintain existing assets, with the remainder going towards expansion for net-zero.)

In total, the cost of government social and climate policies made up £183 (16%) of typical dual-fuel bills in October 2020. This had fallen to £114 (7%) as of the current October 2026 price cap.

Wholesale costs make up the largest share of the current cap, at £757 (44%). Network costs are the second-largest component of bills, at £417 (24%), followed by operating costs at £291 (17%).

Sissons tells Carbon Brief:

“Upgrading the grid is not cost-neutral, but that price tag is dwarfed by the impact of rising gas prices. In fact, net-zero is helping to shelter Britain from the impact of expensive gas. Since the start of the Iran war, electricity bills have begun to decouple from gas bills, thanks to the UK’s increased renewable power and support from contracts for difference.”

“Going further will require British households to have clean electricity, electric vehicles and heat pumps over fossil-fuel cars and boilers. By focusing on growing electricity demand, the government can spread upgrade costs more widely. This can lock in the consistent investment in electrification that is Britain’s best path to stop global turmoil showing up on winter energy bills.”

Media reports suggest the chancellor may take steps to help households and businesses cope with rising energy bills at the upcoming budget on 28 October.

Bills are expected to rise sharply under the January 2027 price cap – potentially by as much as a quarter – in light on ongoing high gas prices.

Related UK budget 2026: Key climate and energy announcements 28.10.2026 UK policy Analysis: How the UK spent more than £11.6bn on climate aid since 2021 28.10.2026 UK policy Q&A: Why are UK energy bills so high – and how could they be lowered? 27.10.2026 Oil and gas Factcheck: UK Conservatives double the ‘cost of net-zero’ after spreadsheet blunder 09.10.2026 UK policy

The post Revealed: Expensive gas behind ‘nearly 90%’ of rise in UK electricity bills appeared first on Carbon Brief.

Categories: I. Climate Science

UNEP: Opportunity has ‘been lost’ to meet 1.5C warming limit without ‘overshoot’

Tue, 10/20/2026 - 22:00
Print article Share

A temporary breach of the 1.5C global-warming limit is now “inevitable” following “decades of insufficient climate action”, says the UN Environment Programme (UNEP).

In its “sobering” annual assessment of the global “emissions gap”, UNEP says that exceeding the Paris Agreement’s 1.5C limit is “imminent and unavoidable”.

This year’s report, titled “The point of return”, says the “task now” is to minimise the size and scale of a temperature “overshoot” and to “establish a credible return path to 1.5C by the end of the century”.

Even the “lowest possible” overshoot scenario for 1.5C would first see global warming peak at around 1.7-1.8C in the middle of the century before declining, says the report.

This scenario – which is “still aligned with the long-term temperature goal of the Paris Agreement” – would require global emissions cuts of 18% by 2030 and 45% by 2035, while ramping up “removals” of carbon dioxide (CO2) from the atmosphere, says UNEP.

In contrast, it notes that the current climate pledges by countries around the world – even if fully implemented – would only reduce emissions by 5-8% by 2030 and 11-15 % by 2035, relative to 2025 levels.

Nonetheless, the report identifies “ample opportunity to accelerate action”, concluding that “countries can still limit peak warming, shorten the duration of overshoot and keep the goals of the Paris Agreement within reach”.

(See Carbon Brief’s detailed coverage of previous reports in 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024 and 2025.)

Article Contents Emissions ‘comparable’ between 2024 and 2025

The report says that emissions in 2025 were “comparable” to those of 2024. 

Global greenhouse gas emissions in 2025 totalled 58.4bn tonnes of CO2 equivalent (GtCO2e), according to the report. This is around 0.16GtCO2e lower than 2024 levels – albeit, with high levels of uncertainty around the numbers. 

In addition, because the figures presented in the 2025 report are “provisional”, the report says, it is “not possible to conclude on the overall direction of emissions in 2025”.

The emissions growth rate has been slowing down in recent decades: the growth rate over 1990-2015 was, on average, 1.7% per year, while emissions have increased by an average of 0.8% per year since 2016.

The chart below shows the total human-driven emissions between 1990 and 2025, separated by greenhouse gas: CO2 from fossil-fuel combustion (dark grey), methane (light grey), nitrous oxide (blue), fluorinated gases (orange) and CO2 from land-use change (green).

Credit: UNEP emissions gap report (2026)

Compared to 2024, emissions from fossil fuels increased slightly, by between 0.4-1.1%, says the report. Emissions of both nitrous oxide and fluorinated gases increased by even higher rates, but since these comprise such a small part of total greenhouse gas emissions, “their contribution to the overall trend is relatively less significant”, it adds. Methane emissions remained steady over that time.

A reduction in the CO2 emissions from land use, land-use change and forestry (LULUCF) “slightly overcompensates” for the rise in emissions of fossil-fuel CO2 and other gases, according to the report. (For more on falling land-use emissions, see Carbon Brief’s explainer.)

The decrease in LULUCF emissions was “partially driven” by a decline in global wildfire emissions, particularly in Africa, South America and parts of south-east Asia. The report notes that these decreases were “likely connected to the ceasing of El Niño conditions”. The ongoing “super” El Niño is likely to cause LULUCF emissions to increase in 2026.

The report also examines countries’ contributions to the increase in greenhouse gas emissions – although here, it excludes LULUCF due to a lag in national-level emissions inventories for the sector. 

Among the top six emitters, only the US saw a notable increase in 2025, says the report, whereas emissions in China, the EU, India, Russia and Indonesia were relatively flat. Emissions in the US grew by 133m tonnes of CO2e (MtCO2e) in 2025, due to increased electricity demand and a “rebound” of coal power. This growth in US emissions is likely due to a combination of market factors, including increased heating demand due to a colder winter, higher gas prices and overall increased electricity demand.

The changes for the six largest emitters, the rest of the G20 countries, international transport and the rest of the world can be seen in the chart below.

Credit: UNEP emissions gap report (2026)

Across most sectors, progress towards decarbonisation has been “modest”, says the report. The most notable decreases have occurred in the power sector and the transport sector, due to the growth of solar power and electric vehicles. The report notes:

“There are emerging signs of transport sector electrification and decarbonisation, but at rates and levels that are not yet sufficient to drive emissions reductions at the global level.”

However, emissions from the transport sector have continued to grow, due to increased demand for international aviation and shipping. 

The report notes that while electricity demand from data centres is “projected to double” by 2030, data centres at present account for less than 1% of total emissions.

New NDCs would reduce emissions ‘far from what is needed’

All countries were meant to submit new UN climate pledges for 2035, known as the third nationally determined contributions (NDCs) to the Paris Agreement, by February 2025. However, 95% of nations missed this deadline.

Many countries came forward with their third NDCs at a UN climate summit in New York in September 2025 and at the COP30 climate summit in Brazil in November 2025.

By the end of COP30, 122 – or 63% – of countries had produced their third NDCs.

The report says that figure had risen to 124 by 1 September 2026, with pledges now covering 75% of global emissions.

This is more than double the number of third NDCs available at the time of the 2025 emissions gap report – allowing a “more robust basis for assessing collective progress and ambition toward 2035”, says the report.

Although the first global stocktake of action towards the goals of the Paris Agreement called for countries to submit targets “covering all greenhouse gases”, there has only been a “modest” increase in the number of nations to do so, according to the report.

It says that eight countries included greenhouse gas targets for the first time in their third NDCs, bringing the total number parties with such targets up to 155.

The third NDCs “increasingly use more robust and verifiable target formulations”, such as standardised baselines, says the report.

However, they have not “meaningfully broadened the scope of sectors and gases covered by the targets”, it adds.

#gform_wrapper_5[data-form-index="0"].gform-theme,[data-parent-form="5_0"]{--gf-color-primary: #204ce5;--gf-color-primary-rgb: 32, 76, 229;--gf-color-primary-contrast: #fff;--gf-color-primary-contrast-rgb: 255, 255, 255;--gf-color-primary-darker: #001AB3;--gf-color-primary-lighter: #527EFF;--gf-color-secondary: #fff;--gf-color-secondary-rgb: 255, 255, 255;--gf-color-secondary-contrast: #112337;--gf-color-secondary-contrast-rgb: 17, 35, 55;--gf-color-secondary-darker: #F5F5F5;--gf-color-secondary-lighter: #FFFFFF;--gf-color-out-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-out-ctrl-light-rgb: 17, 35, 55;--gf-color-out-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-out-ctrl-light-lighter: #F5F5F5;--gf-color-out-ctrl-dark: #585e6a;--gf-color-out-ctrl-dark-rgb: 88, 94, 106;--gf-color-out-ctrl-dark-darker: #112337;--gf-color-out-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-color-in-ctrl: #fff;--gf-color-in-ctrl-rgb: 255, 255, 255;--gf-color-in-ctrl-contrast: #112337;--gf-color-in-ctrl-contrast-rgb: 17, 35, 55;--gf-color-in-ctrl-darker: #F5F5F5;--gf-color-in-ctrl-lighter: #FFFFFF;--gf-color-in-ctrl-primary: #204ce5;--gf-color-in-ctrl-primary-rgb: 32, 76, 229;--gf-color-in-ctrl-primary-contrast: #fff;--gf-color-in-ctrl-primary-contrast-rgb: 255, 255, 255;--gf-color-in-ctrl-primary-darker: #001AB3;--gf-color-in-ctrl-primary-lighter: #527EFF;--gf-color-in-ctrl-light: rgba(17, 35, 55, 0.1);--gf-color-in-ctrl-light-rgb: 17, 35, 55;--gf-color-in-ctrl-light-darker: rgba(104, 110, 119, 0.35);--gf-color-in-ctrl-light-lighter: #F5F5F5;--gf-color-in-ctrl-dark: #585e6a;--gf-color-in-ctrl-dark-rgb: 88, 94, 106;--gf-color-in-ctrl-dark-darker: #112337;--gf-color-in-ctrl-dark-lighter: rgba(17, 35, 55, 0.65);--gf-radius: 3px;--gf-font-size-secondary: 14px;--gf-font-size-tertiary: 13px;--gf-icon-ctrl-number: url("data:image/svg+xml,%3Csvg width='8' height='14' viewBox='0 0 8 14' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M4 0C4.26522 5.96046e-08 4.51957 0.105357 4.70711 0.292893L7.70711 3.29289C8.09763 3.68342 8.09763 4.31658 7.70711 4.70711C7.31658 5.09763 6.68342 5.09763 6.29289 4.70711L4 2.41421L1.70711 4.70711C1.31658 5.09763 0.683417 5.09763 0.292893 4.70711C-0.0976311 4.31658 -0.097631 3.68342 0.292893 3.29289L3.29289 0.292893C3.48043 0.105357 3.73478 0 4 0ZM0.292893 9.29289C0.683417 8.90237 1.31658 8.90237 1.70711 9.29289L4 11.5858L6.29289 9.29289C6.68342 8.90237 7.31658 8.90237 7.70711 9.29289C8.09763 9.68342 8.09763 10.3166 7.70711 10.7071L4.70711 13.7071C4.31658 14.0976 3.68342 14.0976 3.29289 13.7071L0.292893 10.7071C-0.0976311 10.3166 -0.0976311 9.68342 0.292893 9.29289Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-select: url("data:image/svg+xml,%3Csvg width='10' height='6' viewBox='0 0 10 6' fill='none' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath fill-rule='evenodd' clip-rule='evenodd' d='M0.292893 0.292893C0.683417 -0.097631 1.31658 -0.097631 1.70711 0.292893L5 3.58579L8.29289 0.292893C8.68342 -0.0976311 9.31658 -0.0976311 9.70711 0.292893C10.0976 0.683417 10.0976 1.31658 9.70711 1.70711L5.70711 5.70711C5.31658 6.09763 4.68342 6.09763 4.29289 5.70711L0.292893 1.70711C-0.0976311 1.31658 -0.0976311 0.683418 0.292893 0.292893Z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-icon-ctrl-search: url("data:image/svg+xml,%3Csvg width='640' height='640' xmlns='http://www.w3.org/2000/svg'%3E%3Cpath d='M256 128c-70.692 0-128 57.308-128 128 0 70.691 57.308 128 128 128 70.691 0 128-57.309 128-128 0-70.692-57.309-128-128-128zM64 256c0-106.039 85.961-192 192-192s192 85.961 192 192c0 41.466-13.146 79.863-35.498 111.248l154.125 154.125c12.496 12.496 12.496 32.758 0 45.254s-32.758 12.496-45.254 0L367.248 412.502C335.862 434.854 297.467 448 256 448c-106.039 0-192-85.962-192-192z' fill='rgba(17, 35, 55, 0.65)'/%3E%3C/svg%3E");--gf-label-space-y-secondary: var(--gf-label-space-y-md-secondary);--gf-ctrl-border-color: #686e77;--gf-ctrl-size: var(--gf-ctrl-size-md);--gf-ctrl-label-color-primary: #112337;--gf-ctrl-label-color-secondary: #112337;--gf-ctrl-choice-size: var(--gf-ctrl-choice-size-md);--gf-ctrl-checkbox-check-size: var(--gf-ctrl-checkbox-check-size-md);--gf-ctrl-radio-check-size: var(--gf-ctrl-radio-check-size-md);--gf-ctrl-btn-font-size: var(--gf-ctrl-btn-font-size-md);--gf-ctrl-btn-padding-x: var(--gf-ctrl-btn-padding-x-md);--gf-ctrl-btn-size: var(--gf-ctrl-btn-size-md);--gf-ctrl-btn-border-color-secondary: #686e77;--gf-ctrl-file-btn-bg-color-hover: #EBEBEB;--gf-field-img-choice-size: var(--gf-field-img-choice-size-md);--gf-field-img-choice-card-space: var(--gf-field-img-choice-card-space-md);--gf-field-img-choice-check-ind-size: var(--gf-field-img-choice-check-ind-size-md);--gf-field-img-choice-check-ind-icon-size: var(--gf-field-img-choice-check-ind-icon-size-md);--gf-field-pg-steps-number-color: rgba(17, 35, 55, 0.8);}

URLThis field is for validation purposes and should be left unchanged.

Free fortnightly newsletter

.preheader p{ margin-top: 0; font-family: 'PT Sans', sans-serif; font-weight: var(--type--3--font-weight--bold); color: var(--button--color); font-size: var(--button--font-size, inherit); } .newsletter-inline{ display: flex; border: solid 1px #333333; padding: 1em; background: #ffffff; } .inline-email{ display:inline-block; margin-right:1em; margin-top:0 !important; margin-bottom:0.5em; } #field_submit{ display:inline-block; margin-top:0 !important; } .gform_wrapper .gfield+.gfield{ margin-top:0 } Email gform.initializeOnLoaded( function() {gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery('#gform_ajax_frame_5').on('load',function(){var contents = jQuery(this).contents().find('*').html();var is_postback = contents.indexOf('GF_AJAX_POSTBACK') >= 0;if(!is_postback){return;}var form_content = jQuery(this).contents().find('#gform_wrapper_5');var is_confirmation = jQuery(this).contents().find('#gform_confirmation_wrapper_5').length > 0;var is_redirect = contents.indexOf('gformRedirect(){') >= 0;var is_form = form_content.length > 0 && ! is_redirect && ! is_confirmation;var mt = parseInt(jQuery('html').css('margin-top'), 10) + parseInt(jQuery('body').css('margin-top'), 10) + 100;if(is_form){jQuery('#gform_wrapper_5').html(form_content.html());if(form_content.hasClass('gform_validation_error')){jQuery('#gform_wrapper_5').addClass('gform_validation_error');} else {jQuery('#gform_wrapper_5').removeClass('gform_validation_error');}setTimeout( function() { /* delay the scroll by 50 milliseconds to fix a bug in chrome */ jQuery(document).scrollTop(jQuery('#gform_wrapper_5').offset().top - mt); }, 50 );if(window['gformInitDatepicker']) {gformInitDatepicker();}if(window['gformInitPriceFields']) {gformInitPriceFields();}var current_page = jQuery('#gform_source_page_number_5').val();gformInitSpinner( 5, 'http://www.carbonbrief.org/wp-content/plugins/gravityforms/images/spinner.svg', false );jQuery(document).trigger('gform_page_loaded', [5, current_page]);window['gf_submitting_5'] = false;}else if(!is_redirect){var confirmation_content = jQuery(this).contents().find('.GF_AJAX_POSTBACK').html();if(!confirmation_content){confirmation_content = contents;}jQuery('#gform_wrapper_5').replaceWith(confirmation_content);jQuery(document).scrollTop(jQuery('#gf_5').offset().top - mt);jQuery(document).trigger('gform_confirmation_loaded', [5]);window['gf_submitting_5'] = false;wp.a11y.speak(jQuery('#gform_confirmation_message_5').text());}else{jQuery('#gform_5').append(contents);if(window['gformRedirect']) {gformRedirect();}}jQuery(document).trigger("gform_pre_post_render", [{ formId: "5", currentPage: "current_page", abort: function() { this.preventDefault(); } }]); if (event && event.defaultPrevented) { return; } const gformWrapperDiv = document.getElementById( "gform_wrapper_5" ); if ( gformWrapperDiv ) { const visibilitySpan = document.createElement( "span" ); visibilitySpan.id = "gform_visibility_test_5"; gformWrapperDiv.insertAdjacentElement( "afterend", visibilitySpan ); } const visibilityTestDiv = document.getElementById( "gform_visibility_test_5" ); let postRenderFired = false; function triggerPostRender() { if ( postRenderFired ) { return; } postRenderFired = true; gform.core.triggerPostRenderEvents( 5, current_page ); if ( visibilityTestDiv ) { visibilityTestDiv.parentNode.removeChild( visibilityTestDiv ); } } function debounce( func, wait, immediate ) { var timeout; return function() { var context = this, args = arguments; var later = function() { timeout = null; if ( !immediate ) func.apply( context, args ); }; var callNow = immediate && !timeout; clearTimeout( timeout ); timeout = setTimeout( later, wait ); if ( callNow ) func.apply( context, args ); }; } const debouncedTriggerPostRender = debounce( function() { triggerPostRender(); }, 200 ); if ( visibilityTestDiv && visibilityTestDiv.offsetParent === null ) { const observer = new MutationObserver( ( mutations ) => { mutations.forEach( ( mutation ) => { if ( mutation.type === 'attributes' && visibilityTestDiv.offsetParent !== null ) { debouncedTriggerPostRender(); observer.disconnect(); } }); }); observer.observe( document.body, { attributes: true, childList: false, subtree: true, attributeFilter: [ 'style', 'class' ], }); } else { triggerPostRender(); } } );} );

If the third NDCs were implemented in full, it would lower global emissions in 2035 by 3.5GtCO2e, compared to 2030 targets, says the report.

It says this is “far from the global reductions needed to get on track to achieving the long-term temperature goal of the Paris Agreement”.

Compared to the 2025 report, the projections for global emissions in 2035 as a result the third NDCs are about 3GtCO2e higher. 

This is down to the formal withdrawal of the US from the Paris Agreement, as well as from “harmonising” the projections with the latest historical emissions data, says the report.

The global “implementation gap” for 2030 – defined as the difference between countries’ projected emissions under current policies and the emissions expected if they implemented their NDCs in full – has narrowed compared to last year, says the report.

However, this is mainly due to the withdrawal of the US from the Paris Agreement. In other words, the narrowing of the gap is due to “policy changes, not policy progress”.

The implementation gap for both 2030 and 2035 are of a similar size – suggesting “enhanced collective ambition of 2035 NDCs is weak”, continues the report. 

Some 14 of the 17 G20 nations, counting the EU as one, have submitted new NDCs with emissions targets for 2035, says the report. 

However, the emissions trajectories of the majority of G20 nations remain “misaligned” with their net-zero targets, says the report.

The chart below shows the gap between nations’ current emission trajectories and their NDC targets for 2030 and 2035 (diamonds), along with their net-zero targets (circles).

Credit: UNEP emissions gap report (2026)

All G20 members except Japan, the EU and UK would need to “accelerate their emission reduction rates based on current policies” to achieve their net-zero targets, adds the report.

Emissions gap is ‘still large’

The report goes on to estimate the “emissions gap” – the difference between projected global emissions if all NDCs are met in full and what would be needed to achieve the long-term temperature goal of the Paris Agreement.

The report assesses a range of potential future pathways – some that have been assessed in previous reports and some new. The three core “NDC scenarios” are:

  • A “current policy” scenario, which projects global greenhouse gas emissions based only on policies adopted and implemented as of November 2025.
  • An “unconditional NDCs” scenario, which projects greenhouse gas emissions assuming full implementation of the most recent NDCs that do not depend on external support (with a cut-off date of 1 September 2026).
  • A “conditional NDCs” scenario that further assumes that the implementation of NDCs that depend on external support, such as climate finance from wealthier countries (also with a cut-off date of 1 September 2026). 

As with the 2025 report, the authors also analyse two “scenario extensions”, which explore the implications of current policies, NDCs and net-zero pledges beyond 2035:

  • A “current policies continuing” scenario, which “follows current policies to 2035 and assumes a continuation of similar efforts thereafter”.
  • A “conditional NDCs plus all net-zero pledges” scenario, which is “the most optimistic scenario included”. This scenario assumes the “conditional NDC” scenario is achieved until 2035 and then all net-zero or other long-term low emissions developments strategies are followed thereafter, excluding that of the US.

In previous reports, the authors have compared all of these scenarios against pathways that hold warming below 2C and 1.8C, or against 1.5C with no more than 0.1C of “overshoot”.

(Overshoot refers to pathways where global warming first exceeds the 1.5C limit and is then brought back down through techniques that remove carbon from the atmosphere.)

These three warming pathways were drawn from the scenario database for the mitigation report of the Intergovernmental Panel on Climate Change’s (IPCC) sixth assessment (AR6). 

These scenarios were developed around 2015-20 and assumed that the world would pursue immediate emissions cuts from around 2020. However, in reality, emissions have continued to rise. As a result, the report says, those temperature pathways “have become outdated”.

For example, the authors explain, emissions are currently around 80% above the level that would have been needed by 2030 to limit warming to 1.5C with no or “low” overshoot.

Therefore, as a result of “decades of insufficient emission reductions”, the possibility of limiting warming to 1.5C with limited or no overshoot “has now been lost”, say the authors.

The 2026 report instead “draws on a new generation of pathways in which ambitious mitigation action starts from 2025”. These are:

  • A “below 2C” scenario, which is consistent with keeping global warming below 2C throughout the 21st century, with at least a 66% chance.
  • A “lowest 1.5C overshoot” scenario, which keeps warming to 1.7-1.8C throughout the entire century (with a 66% chance) and returns it to 1.5C by 2100 (with a 50% chance). 

These updated scenarios “reflect the failure to cut global emissions to date”, note the authors.

The overshoot scenario “minimise[s] both the magnitude and duration of overshoot above 1.5C”, say the authors, while meeting a defined set of “feasibility and sustainability guardrails”. These include, for example, balancing the demands of bioenergy and large-scale carbon removals with biodiversity and food security needs.

The task to “minimise” overshoot reflects the wording of the “Global Mutirão” package of measures agreed at the COP30 climate summit in Brazil in 2025. This acknowledged – for the first time in a COP text – that there was likely to be an overshoot of 1.5C and urged that both its extent and duration needed to be “limit[ed]”.

The chart below, taken from the report, shows how the emissions pathways compare to each other, as well as historical emissions and the 1.5C pathway used in previous reports.

The right-hand section shows the “unbridgeable gap” to pathways that limit warming to 1.5C with overshoot of no more than 0.1C.

The left-hand chart shows projected emissions under the three NDC scenarios (median estimates), as well as the “below 2C” and “lowest 1.5C overshoot” scenarios (central estimates, plus the full min-max range) out to 2035. The right-hand chart shows the emissions gap in 2035 to the conditional NDCs. Credit: UNEP emissions gap report (2026)

The lowest 1.5C overshoot scenario requires “immediate, deep, rapid and sustained reductions”, says the report. Global greenhouse gas emissions would need to fall by 18% by 2030, 45% by 2035 and 69% by 2040, relative to 2025 levels. Global net-zero CO2 emissions would be reached around 2050.

These benchmarks for 1.5C effectively replace those in the IPCC’s AR6, which specified reductions of 43% by 2030 and 60% by 2035, relative to 2019 levels. These have been frequently cited in COP decisions and other declarations. 

With the ongoing uncertainty around the timeline for the IPCC’s seventh assessment reports, the benchmarks in the UNEP report provide an important update on the emissions cuts required to limit overshoot.

For the 2C scenario, global greenhouse gas emissions would need to drop by 11% by 2030, 24% by 2035 and 37% by 2040, relative to 2025 levels, the report says. 

The table below sets out the estimated emissions from the NDC scenarios and the gaps to the new “below 2C” and “lowest 1.5C overshoot” scenarios.

Credit: UNEP emissions gap report (2026)

These emissions gaps are “still large”, say the authors, with current NDCs and net-zero targets not being “enough to minimise overshoot”. 

(The report also notes that because the 1.5C and 2C scenarios have changed, the emissions gaps in 2026 are not directly comparable with previous reports.)

Putting the world on track for the lowest 1.5C scenario would require global emissions to fall by 5.7% per year over the next decade, it says.

This is “even more rapid” than the 5% annual drop that would have been required to meet the 1.5C limit without overshoot, had emissions cuts started in 2019, adds the report.

Every five years of delay will add 0.1C

To explore different future emission pathways and their impact on global temperatures, the report uses a simple climate model, or “emulator”, called FaIR. 

The report estimates that current national policies would cause 2.8C of global warming over the century, with a 66% chance. Under this scenario, there is also a 10% chance that warming would exceed 3.4C.

If unconditional NDCs are implemented, warming this century could reach 2.7C, with a 66% chance, it says.

If all NDCs are fully implemented, warming could reach 2.6C by 2100, with a 66% chance.

In the “most optimistic” scenario, where countries fully implement their NDCs and reach their net-zero targets, there is a 66% chance that warming could be limited to 2.1C by the end of this century.

These results are illustrated on the chart below, with boxes showing a 10-90% certainty range under a central estimate of future emissions and vertical lines showing the “combined uncertainty across greenhouse gas emissions projections and Earth-system response uncertainty”.

Credit: UNEP emissions gap report (2026)

Compared to last year, the central warming estimates in 2100 for the NDC scenarios are about 0.2-0.3C higher, says the report.

This is because the projected global emissions as a result of the 2035 NDCs have increased. Again, this is down to the US withdrawal from the Paris Agreement and higher estimates of historical emissions.

While fully implementing all NDCs and net-zero pledges could cut warming by 0.6C, this would still be 0.25C higher than the lowest 1.5C overshoot pathways, says the report.

It adds that this “highlights the need both to close the implementation gap between current policies and existing NDCs and net-zero pledges, and to strengthen NDCs or exceed existing commitments”.

As noted above, the report continues that, “with immediate, deep, rapid and sustained emission cuts”, it would still be possible to limit peak warming to 1.7-1.8C, with a 66% chance.

If these emission cuts are “combined with an ambitious and sustainable scale-up of CO2 removal”, then warming could be returned to 1.5C by the end of the century, it continues.

However, “continued delay” in reducing global emissions would “further worsen the climate outcomes still achievable”, it says, adding:

“For every five years of delay in starting to cut emissions, around another 0.1C of warming will be locked in.

“Continued insufficient emission reductions over the next decades could make warming of 2C inevitable, putting the Paris Agreement at risk of failure.”

The figure below shows the likelihood of holding warming to 1.5C, 2C and 3C under scenarios over this century where countries continue with current policies, implement unconditional NDCs, all NDCs and both NDCs and net-zero pledges. 

Credit: UNEP emissions gap report (2026) ‘Uneven and insufficient’ progress towards global stocktake

The report also looks at how the world is measuring up against the mitigation actions set out in the first “global stocktake”, which was agreed at the COP28 climate summit in Dubai, in 2023.

The global stocktake is a key part of the Paris regime that assesses progress towards meeting the goals of the agreement. After each stocktake is completed, countries are supposed to increase their ambition as needed to get on track with their collective targets.

The report looks at nine “mitigation action areas” agreed in the outcome of the first global stocktake. Overall, the report says, progress in those areas is “uneven and insufficient”.

However, it does note that there are “several remarkable examples of progress that exceeded expectations”. 

In particular, the deployment of renewable energy has increased substantially more than was expected by future projections made in 2015 using then-current policies, notes the report. The increase in overall renewable capacity was three times higher than expected under the previous projections. As a result, CO2 emissions related to energy generation increased by 40% less than was expected.

The chart below shows how key energy indicators have changed (black circles), compared to the 2015 projections (white circles).

Source: UNEP emissions gap report (2026)

According to the report, “only renewables are on track for 2030…though continued adoption and implementation of supportive policies will be required to sustain such growth”. 

The world is “lagging particularly far behind” where it needs to be to limit warming to the lowest possible overshoot of the 1.5C target in two areas: transitioning away from fossil fuels and halting deforestation, says the report. (Deforestation is the largest contributor to LULUCF emissions.) 

The report notes that the amount of energy sourced from fossil fuels “continues to rise globally, albeit at a slower rate than was projected in 2015”. Under current policies, it says, global coal usage may peak by 2030, but oil and gas use is “likely to keep increasing”. 

Furthermore, it says, global efforts to stop deforestation by 2030 “remain significantly off track”. Achieving the end-of-decade target would require accelerating the slowdown in forest loss by “more than a factor of 10”. 

The report also includes modelling that shows what level of ambition is needed to achieve the “lowest overshoot” scenario – that which will “minimise the magnitude and duration of overshoot”. This lowest overshoot scenario results in a maximum warming of around 1.7C, with the world returning below 1.5C by 2100, report authors explained in a press briefing.

(It is “no longer feasible”, the report notes, to limit global warming to a temperature rise of 1.5C with an overshoot of no more than 0.1C.)

It provides projections for the ambition needed by 2030, 2035, 2040 and 2050 to achieve this lowest overshoot scenario. For example, minimising overshoot will require reducing the amount of energy derived from coal by 15% by 2030, compared to 2025 levels. Coal energy should be reduced 55% by 2035, 75% by 2040 and 90% by 2050, shows the report.

For oil and gas, smaller reductions are needed in the short term – 5% and 0%, respectively, by 2030 – but with steep declines in the following years. By 2050, energy from oil should be reduced by 65% and energy from gas should be reduced by 75%, says the report.

The chart below shows the outcomes needed in several energy- and electricity-related areas to limit the peak and duration of overshoot. 

The black line shows the observed changes from 2020 to 2025, while the dashed black line shows 2015 projections based on then-current policies. The thick blue line shows the “minimum collective effort” needed to achieve the lowest overshoot scenario.

Source: UNEP emissions gap report (2026)

(There are a range of different ways to reach the “lowest overshoot” scenario. The “minimum collective effort” illustrates the lowest contribution from each sector or fuel that would still be compatible with reaching the “lowest overshoot” scenario overall.)

Article information

Dr Simon Evans, Carbon Brief’s deputy editor and senior policy editor, is an observer of the UNEP gap report’s steering committee.

UNEP (2026), Emissions gap report 2026: The point of return: Urgent emissions cuts needed to come back from overshoot of 1.5C, doi: https://:unep.org/en/publications/report/emissions-gap-report-2026

Related Guest post: Why tough methane cuts are crucial for keeping warming ‘well-below’ 2C 01.09.2026 Climate pollutants Why land-use emissions have fallen by a third this century – in six charts 17.08.2026 Climate pollutants Experts: Why carbon removal needs a ‘major scale up’ to return warming to 1.5C 19.06.2026 Carbon dioxide removal Analysis: What are the causes of recent record-high global temperatures? 10.12.2025 Climate pollutants

The post UNEP: Opportunity has ‘been lost’ to meet 1.5C warming limit without ‘overshoot’ appeared first on Carbon Brief.

Categories: I. Climate Science

Pages

The Fine Print I:

Disclaimer: The views expressed on this site are not the official position of the IWW (or even the IWW’s EUC) unless otherwise indicated and do not necessarily represent the views of anyone but the author’s, nor should it be assumed that any of these authors automatically support the IWW or endorse any of its positions.

Further: the inclusion of a link on our site (other than the link to the main IWW site) does not imply endorsement by or an alliance with the IWW. These sites have been chosen by our members due to their perceived relevance to the IWW EUC and are included here for informational purposes only. If you have any suggestions or comments on any of the links included (or not included) above, please contact us.

The Fine Print II:

Fair Use Notice: The material on this site is provided for educational and informational purposes. It may contain copyrighted material the use of which has not always been specifically authorized by the copyright owner. It is being made available in an effort to advance the understanding of scientific, environmental, economic, social justice and human rights issues etc.

It is believed that this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have an interest in using the included information for research and educational purposes. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner. The information on this site does not constitute legal or technical advice.