Assessment of cloud and associated radiation fields from a GAN stochastic cloud subcolumn generator

TL;DR

This paper introduces a novel machine learning-based stochastic cloud subcolumn generator using CVAE-GAN and U-Net architectures, improving cloud overlap representation and radiative transfer accuracy in Earth System Models.

Contribution

The study develops a new ML generator trained on satellite data that better captures cloud variability and reduces biases compared to traditional methods.

Findings

Accurately reproduces bimodal cloud overlap distributions.

Halves the RMSE in cloud-top pressure and optical thickness histograms.

Reduces global-mean shortwave cloud radiative effect bias by a factor of three.

Abstract

Modern Earth System Models (ESMs) operate on horizontal scales far larger than typical cloud features, requiring stochastic subcolumn generators to represent subgrid horizontal and vertical cloud variability. Traditional physically-based generators often rely on analytical cloud overlap paradigms, such as exponential-random decorrelation, which can struggle to capture the complex, anti-correlated behavior of non-contiguous cloud layers. In this study, we introduce a novel two-stage machine learning subcolumn generator for the GEOS atmospheric model, utilizing a Conditional Variational Autoencoder combined with a Generative Adversarial Network (CVAE-GAN) and a U-Net architecture. Trained on a merged CloudSat-CALIPSO height-resolved cloud optical depth dataset, the ML generator creates 56 stochastic subcolumns representing cloud occurrence and optical depth profiles. Evaluated against the established R\"{a}is\"{a}nen, the ML approach accurately reproduces bimodal cloud overlap distributions, significantly reduces biases in grid-mean statistics, and halves the root-mean-square error in ISCCP-style cloud-top pressure and optical thickness joint histograms. The improvements brought by our deep generative models translate into more accurate offline radiative transfer calculations, reducing the global-mean shortwave top-of-atmosphere cloud radiative effect bias by a factor of three. Provided that the generator can be accelerated on CPUs, this offers a practical pathway to reduce structural errors at the cloud-radiation interface.