CloudCast: A Satellite-Based Dataset and Baseline for Forecasting Clouds

TL;DR

This paper introduces CloudCast, a comprehensive satellite-based dataset with high-resolution cloud images and annotations, enabling machine learning research for cloud forecasting, which is crucial for improving weather prediction accuracy.

Contribution

The paper presents the first publicly available high-resolution, global-scale cloud dataset with detailed annotations and evaluates state-of-the-art video prediction models on it.

Findings

Promising results with current models, but significant room for improvement.

Provides a new benchmark for future cloud forecasting research.

Enables machine learning approaches to a previously data-scarce problem.

Abstract

Forecasting the formation and development of clouds is a central element of modern weather forecasting systems. Incorrect clouds forecasts can lead to major uncertainty in the overall accuracy of weather forecasts due to their intrinsic role in the Earth's climate system. Few studies have tackled this challenging problem from a machine learning point-of-view due to a shortage of high-resolution datasets with many historical observations globally. In this paper, we present a novel satellite-based dataset called ``CloudCast''. It consists of 70,080 images with 10 different cloud types for multiple layers of the atmosphere annotated on a pixel level. The spatial resolution of the dataset is 928 x 1530 pixels (3x3 km per pixel) with 15-min intervals between frames for the period 2017-01-01 to 2018-12-31. All frames are centered and projected over Europe. To supplement the dataset, we conduct an evaluation study with current state-of-the-art video prediction methods such as convolutional long short-term memory networks, generative adversarial networks, and optical flow-based extrapolation methods. As the evaluation of video prediction is difficult in practice, we aim for a thorough evaluation in the spatial and temporal domain. Our benchmark models show promising results but with ample room for improvement. This is the first publicly available global-scale dataset with high-resolution cloud types on a high temporal granularity to the authors' best knowledge.