Cloud Classification with Unsupervised Deep Learning

Takuya Kurihana, Ian Foster, Rebecca Willett, Sydney Jenkins, Kathryn Koenig, Ruby Werman, Ricardo Barros Lourenco, Casper Neo, Elisabeth Moyer

arXiv preprint · 30 September 2022 · Argonne National Laboratory, University of California, Berkeley, University of Chicago

Paper PDF

In one sentence. Unsupervised deep learning on MODIS radiances learns cloud features and produces physically meaningful cloud classes without predefined categories.

Abstract

We present a framework for cloud characterization that leverages modern unsupervised deep learning technologies. While previous neural network-based cloud classification models have used supervised learning methods, unsupervised learning allows us to avoid restricting the model to artificial categories based on historical cloud classification schemes and enables the discovery of novel, more detailed classifications. Our framework learns cloud features directly from radiance data produced by NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) satellite instrument, deriving cloud characteristics from millions of images without relying on pre-defined cloud types during the training process. We present preliminary results showing that our method extracts physically relevant information from radiance data and produces meaningful cloud classes.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.