Assertion Supported AI assessment confidence: 92% certainty 4/5 debate potential 1/5

Open Climate Fix halved National Grid's one-hour-ahead solar forecast error

Dr. Jack Kelly · AI in the real world: Solar forecasting, EVs, and virtual power plants · Nov 16, 2023 · at 8:44

Dr. Jack Kelly discusses the performance of Open Climate Fix's machine learning solar forecasting models deployed for the UK's National Grid ESO.

0:00 / 0:09exact quote · 9.1s

720p mp4 · rendered on demand · StarZero watermark

“We halved the error an hour ahead. At 24 hours ahead, I think the error was about 15% lower than their existing forecast.”

quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →

The claim, plainly

Open Climate Fix halved solar power forecasting error one hour ahead and lowered it by about 15% at 24 hours ahead compared to National Grid ESO's existing forecast.

a quote alone can be ambiguous, so this restates it with the references and context of the conversation around it resolved. AI-written; the tape below the player is the source.

Assessment

The claim makes two specific assertions regarding Open Climate Fix's (OCF) performance against National Grid ESO's baseline solar forecast: halving error (a ~50% reduction) at a 1-hour horizon, and reducing error by approximately 15% at a 24-hour horizon. Published documentation, including the State of AI Report 2021 (Slide 100) evaluating the OCF and National Grid ESO deployment, confirms that OCF's model more than halved the Mean Absolute Error (MAE) at a 1-hour lead time and reduced MAE at a 24-hour lead time by 14% (matching 'about 15%'). Subsequent case studies and NIA project reports from National Grid ESO / NESO further corroborate that short-term solar forecast errors were halved.

State of AI Report 2021 (Slide 100) documented that OCF's model operational with National Grid ESO more than halved the MAE at a 1-hour lead time and reduced 24-hour lead time MAE by 14%.
Open Climate Fix and NESO case studies confirmed that the Quartz Solar forecast halved solar forecast errors at short lead times for grid control room operations.
National Grid ESO NIA Project reports (NIA2_NGESO002) validated that OCF machine learning forecasting tools outperformed standard legacy baselines by over 30% to 40% across various operational horizons.

Checked against

[1] medium.com [2] medium.com [3] neso.energy [4] neso.energy [5] openclimatefix.org [6] whiterose.ac.uk [7] openclimatefix.org [8] openclimatefix.org

Open Climate Fix National Grid ESO solar forecasting error halved one hour 15% site:openclimatefix.org "NESO" OR "National Grid" "case study" "State of AI Report 2021" "Open Climate Fix" "Temporal Fusion Transformer" Open Climate Fix National Grid ESO solar PV forecast MAE RMSE benchmark report "Solar PV Nowcasting" "National Grid ESO" "1-hour" "24-hour" error

assessed as of 2026-09-02

This assessment is an AI reading of a public statement, made against the cited references with the confidence shown. It can be wrong. The quote, the tape and the references are all on this page; if something is off, use report an error below. full disclaimer →

Mentioned

National Grid ESO Open Climate Fix

⚑ Spot an error? Report it

More from Dr. Jack Kelly

Insight

Forecast accuracy ranks surprisingly low for transmission system operators

“The skill of the forecast is surprisingly low on the list of priorities when transmission system operators are buying forecasts. So it's more about how understandable is it? How reliable are you? Do you have a track record? Can you deliver the forecast in this…”

Dr. Jack Kelly Nov 16, 2023 ▶ 12:29 AI in the real world: Solar forecasting, EVs, and virtual power plants

Assertion Supported

Kelly: Solar was National Grid ESO's biggest demand forecasting uncertainty

“The biggest source of uncertainty for their demand forecast came from their solar power forecast.”

Dr. Jack Kelly Nov 16, 2023 ▶ 6:56 AI in the real world: Solar forecasting, EVs, and virtual power plants

Insight

Kelly: Numerical weather models struggle to generate detailed cloud maps

“Which is all to say that NWPs, whilst they do a great job of predicting things like humidity and temperature, converting that to a really detailed map of where the clouds are is known to be fairly imperfect at the moment.”

Dr. Jack Kelly Nov 16, 2023 ▶ 8:03 AI in the real world: Solar forecasting, EVs, and virtual power plants

Open Climate Fix halved National Grid's one-hour-ahead solar forecast error

More from Dr. Jack Kelly

Turn any episode into a week of clips.