Subseasonal climate prediction in the western US using Bayesian spatial models

Vishwak Srinivasan, Justin Khim, Arindam Banerjee, Pradeep Ravikumar
Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence, PMLR 161:961-970, 2021.

Abstract

Subseasonal climate forecasting is the task of predicting climate variables, such as temperature and precipitation, in a two-week to two-month time horizon. The primary predictors for such prediction problem are spatio-temporal satellite and ground measurements of a variety of climate variables in the atmosphere, ocean, and land, which however have rather limited predictive signal at the subseasonal time horizon. We propose a carefully constructed spatial hierarchical Bayesian regression model that makes use of the inherent spatial structure of the subseasonal climate prediction task. We use our Bayesian model to then derive decision-theoretically optimal point estimates with respect to various performance measures of interest to climate science. As we show, our approach handily improves on various off-the-shelf ML baselines. Since our method is based on a Bayesian framework, we are also able to quantify the uncertainty in our predictions, which is particularly crucial for difficult tasks such as the subseasonal prediction, where we expect any model to have considerable uncertainty at different test locations under different scenarios.

Cite this Paper

BibTeX
@InProceedings{pmlr-v161-srinivasan21b, title = {Subseasonal climate prediction in the western US using Bayesian spatial models}, author = {Srinivasan, Vishwak and Khim, Justin and Banerjee, Arindam and Ravikumar, Pradeep}, booktitle = {Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence}, pages = {961--970}, year = {2021}, editor = {de Campos, Cassio and Maathuis, Marloes H.}, volume = {161}, series = {Proceedings of Machine Learning Research}, month = {27--30 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v161/srinivasan21b/srinivasan21b.pdf}, url = {https://proceedings.mlr.press/v161/srinivasan21b.html}, abstract = {Subseasonal climate forecasting is the task of predicting climate variables, such as temperature and precipitation, in a two-week to two-month time horizon. The primary predictors for such prediction problem are spatio-temporal satellite and ground measurements of a variety of climate variables in the atmosphere, ocean, and land, which however have rather limited predictive signal at the subseasonal time horizon. We propose a carefully constructed spatial hierarchical Bayesian regression model that makes use of the inherent spatial structure of the subseasonal climate prediction task. We use our Bayesian model to then derive decision-theoretically optimal point estimates with respect to various performance measures of interest to climate science. As we show, our approach handily improves on various off-the-shelf ML baselines. Since our method is based on a Bayesian framework, we are also able to quantify the uncertainty in our predictions, which is particularly crucial for difficult tasks such as the subseasonal prediction, where we expect any model to have considerable uncertainty at different test locations under different scenarios.} }
Endnote
%0 Conference Paper %T Subseasonal climate prediction in the western US using Bayesian spatial models %A Vishwak Srinivasan %A Justin Khim %A Arindam Banerjee %A Pradeep Ravikumar %B Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2021 %E Cassio de Campos %E Marloes H. Maathuis %F pmlr-v161-srinivasan21b %I PMLR %P 961--970 %U https://proceedings.mlr.press/v161/srinivasan21b.html %V 161 %X Subseasonal climate forecasting is the task of predicting climate variables, such as temperature and precipitation, in a two-week to two-month time horizon. The primary predictors for such prediction problem are spatio-temporal satellite and ground measurements of a variety of climate variables in the atmosphere, ocean, and land, which however have rather limited predictive signal at the subseasonal time horizon. We propose a carefully constructed spatial hierarchical Bayesian regression model that makes use of the inherent spatial structure of the subseasonal climate prediction task. We use our Bayesian model to then derive decision-theoretically optimal point estimates with respect to various performance measures of interest to climate science. As we show, our approach handily improves on various off-the-shelf ML baselines. Since our method is based on a Bayesian framework, we are also able to quantify the uncertainty in our predictions, which is particularly crucial for difficult tasks such as the subseasonal prediction, where we expect any model to have considerable uncertainty at different test locations under different scenarios.
APA
Srinivasan, V., Khim, J., Banerjee, A. & Ravikumar, P.. (2021). Subseasonal climate prediction in the western US using Bayesian spatial models. Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 161:961-970 Available from https://proceedings.mlr.press/v161/srinivasan21b.html.

Related Material