Diffusion-LAM: Probabilistic Limited Area Weather Forecasting with Diffusion

TL;DR

Experimental results on the MEPS limited area dataset demonstrate the potential of Diffusion-LAM to deliver accurate probabilistic forecasts, highlighting its promise for limited-area weather prediction.

Abstract

Machine learning methods have been shown to be effective for weather forecasting, based on the speed and accuracy compared to traditional numerical models. While early efforts primarily concentrated on deterministic predictions, the field has increasingly shifted toward probabilistic forecasting to better capture the forecast uncertainty. Most machine learning-based models have been designed for global-scale predictions, with only limited work targeting regional or limited area forecasting, which allows more specialized and flexible modeling for specific locations. This work introduces Diffusion-LAM, a probabilistic limited area weather model leveraging conditional diffusion. By conditioning on boundary data from surrounding regions, our approach generates forecasts within a defined area. Experimental results on the MEPS limited area dataset demonstrate the potential of Diffusion-LAM to deliver accurate probabilistic forecasts, highlighting its promise for limited-area weather prediction.

Figures (25)

• Figure 1: An overview of the forecasting process showing the inputs and outputs of the model.
• Figure 2: The noise process for r_2 (relative humidity). We only show 10 diffusion steps to make the visualization simpler, but in practice use 20 steps when sampling new trajectories.
• Figure 3: Forecasts at 57h lead time for r_2. The faded area constitutes the boundary region. Note the difference in fine-scale details and the consistency with the boundary in the ensemble members.
• Figure 4: The mean of the normalized RMSE, CRPS, and SSR for all variables.
• Figure 5: The interior and boundary of a weather state in our limited area model. The faded area is the 10 outermost grid positions, which we use as the boundary area.