Towards Climate Variable Prediction with Conditioned Spatio-Temporal Normalizing Flows

Christina Winkler, David Rolnick
11/12/2023
cs.LGcs.AI

Official code: cool-xuan/msflow (87 stars, last commit 2024-03-08).

Abstract

This study investigates how conditional normalizing flows can be applied to remote sensing data products in climate science for spatio-temporal prediction. The method is chosen due to its desired properties such as exact likelihood computation, predictive uncertainty estimation and efficient inference and sampling which facilitates faster exploration of climate scenarios. Experimental findings reveal that the conditioned spatio-temporal flow surpasses both deterministic and stochastic baselines in prolonged rollout scenarios. It exhibits stable extrapolation beyond the training time horizon for extended rollout durations. These findings contribute valuable insights to the field of spatio-temporal modeling, with potential applications spanning diverse scientific disciplines.

View on arXivView PDF

Code Implementations(1)

cool-xuan/msflowOfficial65%

The official code for "MSFlow: Multi-Scale Normalizing Flows for Unsupervised Anomaly Detection"

8717PythonJun 22, 20222 years agoNOASSERTION

Similar Papers

Climate Variable Downscaling with Conditional Normalizing Flows

85% match2024

CoDiCast: Conditional Diffusion Model for Global Weather Prediction with Uncertainty Quantification

81% match1 repo102024

Calibrating Geophysical Predictions under Constrained Probabilistic Distributions

81% match2025

PreDiff: Precipitation Nowcasting with Latent Diffusion Models

80% match1 repo1552023

A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation

80% match1 repo372024

ClimODE: Climate and Weather Forecasting with Physics-informed Neural ODEs

80% match2 repos1282026

Cite this paper

@article{winkler2023towards,
  title  = {Towards Climate Variable Prediction with Conditioned Spatio-Temporal Normalizing Flows},
  author = {Christina Winkler and David Rolnick},
  year   = {2023},
  eprint = {2311.06958},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2311.06958v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2311.06958.svg)](https://codeofpaper.com/paper/2311.06958)

Discussion