Generating physically-consistent high-resolution climate data with hard-constrained neural networks

  title={Generating physically-consistent high-resolution climate data with hard-constrained neural networks},
  author={Paula Harder and Qidong Yang and Venkatesh Ramesh and Prasanna Sattigeri and Alex Hern{\'a}ndez-Garc{\'i}a and Campbell D. Watson and Daniel Szwarcman and David Rolnick},
The availability of reliable, high-resolution climate and weather data is important to inform long-term decisions on climate adaptation and mitigation and to guide rapid re- sponses to extreme events. Forecasting models are limited by computational costs and therefore often predict quantities at a coarse spatial resolution. Statistical downscaling can provide an efficient method of upsampling low-resolution data. In this field, deep learning has been applied successfully, often using methods from… 

Figures and Tables from this paper



Spatio-Temporal Downscaling of Climate Data Using Convolutional and Error-Predicting Neural Networks

Numerical weather and climate simulations nowadays produce terabytes of data, and the data volume continues to increase rapidly since an increase in resolution greatly benefits the simulation of

DeepSD: Generating High Resolution Climate Change Projections through Single Image Super-Resolution

DeepSD is presented, a generalized stacked super resolution convolutional neural network (SRCNN) framework for statistical downscaling of climate variables and augments SRCNN with multi-scale input channels to maximize predictability in statisticalDownscaling.

Augmented Convolutional LSTMs for Generation of High-Resolution Climate Change Projections

This work presents auxiliary variables informed spatio-temporal neural architecture for statistical downscaling of precipitation variable over one of the most climatically diversified countries, India and showcases significant improvement gain against two popular state-of-the-art baselines with better ability to predict statistics of extreme events.

Downscaling atmospheric chemistry simulations with physically consistent deep learning

Abstract. Recent advances in deep convolutional neural network (CNN)-based super resolution can be used to downscale atmospheric chemistry simulations with substantially higher accuracy than

Fast and accurate learned multiresolution dynamical downscaling for precipitation

A neural network-based approach for emulating high-resolution modeled precipitation data with comparable statistical properties but at greatly reduced computational cost and results show that the SR technique produces results similar to those of the bilinear interpolator with smoother spatial and temporal distributions and smaller data variabilities and extremes than the high resolution simulations.

ClimAlign: Unsupervised statistical downscaling of climate variables via normalizing flows

ClimAlign is presented, a novel method for unsupervised, generative downscaling using adaptations of recent work in normalizing flows for variational inference, and one of very few proposed methods that allows for efficient sampling of synthetic data.

Adversarial super-resolution of climatological wind and solar data

An adversarial deep learning approach to super resolve wind velocity and solar irradiance outputs from global climate models to scales sufficient for renewable energy resource assessment is introduced and a super-resolution study of renewable energy resources based on climate scenario data from the Intergovernmental Panel on Climate Change’s Fifth Assessment Report is concluded.

CliGAN: A Structurally Sensitive Convolutional Neural Network Model for Statistical Downscaling of Precipitation from Multi-Model Ensembles

A novel downscaling method using a Generative Adversarial Network (GAN), CliGAN, which can downscale large-scale annual maximum precipitation given by simulation of multiple atmosphere-ocean global climate models (AOGCM) from Coupled Model Inter-comparison Project 6 (CMIP6) to regional-level gridded annualmaximum precipitation data.

Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields With a Generative Adversarial Network

A recurrent, stochastic super-resolution GAN that can generate ensembles of time-evolving high-resolution atmospheric fields for an input consisting of a low-resolution sequence of images of the same field is introduced.

Physics-informed learning of aerosol microphysics

This work aims to use machine learning to emulate the microphysics model at sufficient accuracy and reduce the computational cost by being fast at inference time and achieves a speed-up of up to over 64 times faster when compared to the original model.