Efficient modeling of sub-kilometer surface wind with Gaussian processes and neural networks

TL;DR

This paper introduces a novel statistical approach combining Gaussian processes and neural networks to model sub-kilometer surface wind gusts, improving probabilistic predictions by capturing multivariate covariance and integrating diverse data sources.

Contribution

The paper presents a new integrated modeling framework that leverages Gaussian processes and neural networks to enhance surface wind predictions at fine spatial scales.

Findings

Modeling multivariate covariance improves prediction accuracy.

Incorporating multiple data sources enhances probabilistic calibration.

Scalable techniques enable realistic and calibrated wind field generation.

Abstract

Accurately representing surface weather at the sub-kilometer scale is crucial for optimal decision-making in a wide range of applications. This motivates the use of statistical techniques to provide accurate and calibrated probabilistic predictions at a lower cost compared to numerical simulations. Wind represents a particularly challenging variable to model due to its high spatial and temporal variability. This paper presents a novel approach that integrates Gaussian processes and neural networks to model surface wind gusts at sub-kilometer resolution, leveraging multiple data sources, including numerical weather prediction models, topographical descriptors, and in-situ measurements. Results demonstrate the added value of modeling the multivariate covariance structure of the variable of interest, as opposed to only applying a univariate probabilistic regression approach. Modeling the covariance enables the optimal integration of observed measurements from ground stations, which is shown to reduce the continuous ranked probability score compared to the baseline. Moreover, it allows the generation of realistic fields that are also marginally calibrated, aided by scalable techniques such as random Fourier features and pathwise conditioning. We discuss the effect of different modeling choices, as well as different degrees of approximation, and present our results for a case study.