Assertion Supported AI assessment confidence: 95% certainty 4/5 debate potential 1/5

Woollacott: Traditional weather simulations are constrained to one-kilometer resolution

Emma Woollacott · AI in the real world: Solar forecasting, EVs, and virtual power plants · Nov 16, 2023 · at 5:18

Emma Woollacott discusses the computational limitations of conventional supercomputer-based numerical weather prediction models used for grid management.

0:00 / 0:20exact quote · 20.4s

720p mp4 · rendered on demand · StarZero watermark

“To do this properly, thoroughly it means an awful lot of computing power is required, and an awful lot of time, so the resolution can't be particularly high, we're looking at about a kilometre square, and they can't be particularly fast, because there's a huge amount of number crunching to be done.”

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

Traditional numerical weather prediction simulations require massive computing power and time, limiting their spatial resolution to approximately one square kilometer.

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

Operational meteorological literature and supercomputing benchmarks circa 2023 confirm that state-of-the-art high-resolution numerical weather prediction (NWP) models are constrained by computational power and real-time operational delivery deadlines to horizontal grid resolutions of approximately 1 km to 3 km (1 to 9 square kilometers). Halving grid spacing increases computational costs approximately eightfold due to 3D grid cell scaling and Courant–Friedrichs–Lewy (CFL) time-stepping constraints. Operational regional models such as NOAA HRRR (3 km), UK Met Office UKV (1.5 km), Météo-France AROME (1.3 km), and specialized commercial suites (e.g., Meteomatics, Spire at 1 km) reflect these practical physical and computational limits.

Operational regional NWP systems circa 2023 (e.g., NOAA HRRR at 3 km, UK Met Office UKV at 1.5 km, Météo-France AROME-France at 1.3 km, and commercial high-resolution runs at 1 km) operate at grid spacings bounded around 1 to 3 km.
Meteorological scaling literature confirms that halving horizontal grid spacing requires ~8x computing power due to 4x cell count multiplication combined with halving the dynamic integration time step to ensure numerical stability under the CFL condition.
Strict operational forecast delivery windows (typically requiring runs to finish within 30–60 minutes) constrain operational grid resolution to these bounds.

Checked against

[1] wikipedia.org [2] github.io [3] infoplaza.com [4] medium.com [5] gribstream.com [6] wikipedia.org [7] researchgate.net [8] noaa.gov

operational numerical weather prediction horizontal resolution km circa 2023 NOAA HRRR horizontal grid spacing resolution km computational limits numerical weather prediction spatial resolution 1 km operational NWP regional model resolution AROME ICON-D2 UKV km "AROME" "Meteo-France" operational horizontal resolution km

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

numerical weather forecasts supercomputer

⚑ Spot an error? Report it

More from Emma Woollacott

Assertion Partly supported

UK ran for 47 consecutive days without coal power before heatwave

“47 days the UK had been without needing any coal plants.”

Emma Woollacott Nov 16, 2023 ▶ 3:15 AI in the real world: Solar forecasting, EVs, and virtual power plants

Woollacott: Traditional weather simulations are constrained to one-kilometer resolution

More from Emma Woollacott

Turn any episode into a week of clips.