A Generative Deep Learning Approach to Stochastic Downscaling of Precipitation Forecasts

Abstract

Despite continuous improvements, precipitation forecasts are still not as accurate and reliable as those of other meteorological variables. A major contributing factor to this is that several key processes affecting precipitation distribution and intensity occur below the resolved scale of global weather models. Generative adversarial networks (GANs) have been demonstrated by the computer vision community to be successful at super-resolution problems, i.e., learning to add fine-scale structure to coarse images. Leinonen et al. (2020) previously applied a GAN to produce ensembles of reconstructed high-resolution atmospheric fields, given coarsened input data. In this paper, we demonstrate this approach can be extended to the more challenging problem of increasing the accuracy and resolution of comparatively low-resolution input from a weather forecasting model, using high-resolution radar measurements as a "ground truth". The neural network must learn to add resolution and structure whilst accounting for non-negligible forecast error. We show that GANs and VAE-GANs can match the statistical properties of state-of-the-art pointwise post-processing methods whilst creating high-resolution, spatially coherent precipitation maps. Our model compares favourably to the best existing downscaling methods in both pixel-wise and pooled CRPS scores, power spectrum information and rank histograms (used to assess calibration). We test our models and show that they perform in a range of scenarios, including heavy rainfall.

Revised version 28/7/22

References in corpus (7)

• Conditional Generative Adversarial Nets
• Machine Learning for Precipitation Nowcasting from Radar Images
• A Generative Deep Learning Approach to Stochastic Downscaling of Precipitation Forecasts
• Probabilistic quantitative precipitation field forecasting using a two-stage spatial model
• Increasing the accuracy and resolution of precipitation forecasts using deep generative models
• MS-nowcasting: Operational Precipitation Nowcasting with Convolutional LSTMs at Microsoft Weather
• Investigating two super-resolution methods for downscaling precipitation: ESRGAN and CAR

Cited by in corpus (11)

• A Generative Deep Learning Approach to Stochastic Downscaling of Precipitation Forecasts
• On the modern deep learning approaches for precipitation downscaling
• Machine learning applications for weather and climate need greater focus on extremes
• Generative machine learning methods for multivariate ensemble post-processing
• Can AI be enabled to dynamical downscaling? A Latent Diffusion Model to mimic km-scale COSMO5.0\_CLM9 simulations
• Global spatio-temporal downscaling of ERA5 precipitation through generative AI
• On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates
• On the Effectiveness of Neural Operators at Zero-Shot Weather Downscaling
• Physics-constrained generative machine learning-based high-resolution downscaling of Greenland's surface mass balance and surface temperature
• Downscaling Extreme Precipitation with Wasserstein Regularized Diffusion
• Efficient Kilometer-Scale Precipitation Downscaling with Conditional Wavelet Diffusion