Lightning-Fast Thunderstorm Warnings: Predicting Severe Convective Environments with Global Neural Weather Models

Feldmann, Monika; Beucler, Tom; Gomez, Milton; Martius, Olivia (17 June 2024). Lightning-Fast Thunderstorm Warnings: Predicting Severe Convective Environments with Global Neural Weather Models (arXiv). Cornell University

Text 2406.09474v1.pdf - Published Version Restricted to registered users only Available under License Publisher holds Copyright. Download (32MB)

The recently released suite of neural weather models can produce multi-day, medium-range forecasts within seconds, with a skill on par with state-of-the-art operational forecasts. Traditional neural model evaluation predominantly targets global scores on single
levels. Specific prediction tasks, such as severe convective environments, require much
more precision on a local scale and with the correct vertical gradients in between levels.
With a focus on the convective season of global hotspots in 2020, we assess the skill of
three top-performing neural weather models (Pangu-Weather, GraphCast, FourCastNet)
for two variables used to assess convective activity (Convective Available Potential Energy - CAPE, and Deep Layer Shear - DLS) at lead-times of up to 10 days against the
ERA-5 reanalysis and the IFS operational numerical weather prediction model. Looking
at the example of a US tornado outbreak on April 12 and 13, 2020, all models predict
elevated CAPE and DLS values multiple days in advance. The spatial structures in the
neural weather models are smoothed in comparison to IFS and ERA-5. The models show
differing biases in the prediction of CAPE values, with GraphCast capturing the value
distribution the most accurately and FourCastNet showing a consistent underestimation.
In seasonal analyses around the globe, we generally see the highest performance by Graph Cast and Pangu-Weather, which match or even exceed the performance of IFS. CAPE
derived from vertically coarse pressure levels of neural weather models nonetheless lacks
the precision of CAPE derived from the vertically fine resolution of numerical models.
The promising results here indicate that a direct prediction of CAPE in neural weather
models is likely to be skillful. This would open unprecedented opportunities for fast and
inexpensive predictions of severe weather phenomena. By advancing the assessment of
large neural weather models towards process-based evaluations we lay the foundation for
hazard-driven applications of artificial-intelligence-based weather forecasts.

Interest & Impact

Downloads

3 since deposited on 07 Aug 2024
3 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item