Deep Learning for Bias‐Correcting CMIP6‐Class Earth System Models

Philipp Hess; Stefan Lange; Christof Schötz; Niklas Boers

doi:10.1029/2023ef004002

journal article Open Access Oct 01, 2023

Deep Learning for Bias‐Correcting CMIP6‐Class Earth System Models

Philipp Hess

Stefan Lange

Christof Schötz

Niklas Boers

Earth's Future Vol. 11 No. 10 · American Geophysical Union (AGU)

View at Publisher Save 10.1029/2023ef004002

Abstract

AbstractThe accurate representation of precipitation in Earth system models (ESMs) is crucial for reliable projections of the ecological and socioeconomic impacts in response to anthropogenic global warming. The complex cross‐scale interactions of processes that produce precipitation are challenging to model, however, inducing potentially strong biases in ESM fields, especially regarding extremes. State‐of‐the‐art bias correction methods only address errors in the simulated frequency distributions locally at every individual grid cell. Improving unrealistic spatial patterns of the ESM output, which would require spatial context, has not been possible so far. Here, we show that a postprocessing method based on physically constrained generative adversarial networks (cGANs) can correct biases of a state‐of‐the‐art, CMIP6‐class ESM both in local frequency distributions and in the spatial patterns at once. While our method improves local frequency distributions equally well as gold‐standard bias‐adjustment frameworks, it strongly outperforms any existing methods in the correction of spatial patterns, especially in terms of the characteristic spatial intermittency of precipitation extremes.

Topics

No keywords indexed for this article. Browse by subject →

References

52

[1]

10.1175/bams-d-17-0138.1

[2]

Accurate medium-range global weather forecasting with 3D neural networks

Kaifeng Bi, Lingxi Xie, Hengheng Zhang et al.

Nature 10.1038/s41586-023-06185-3

[3]

10.1029/2010jd014262

[4]

Multivariate quantile mapping bias correction: an N-dimensional probability density function transform for climate model simulations of multiple variables

Alex J. Cannon

Climate Dynamics 10.1007/s00382-017-3580-6

[5]

Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes?

Alex J. Cannon, Stephen R. Sobie, Trevor Q. Murdock

Journal of Climate 10.1175/jcli-d-14-00754.1

[6]

Cmip6 gfdl‐esm4 model data. (n.d.).Cmip6 gfdl‐esm4 model data[Dataset]. Retrieved fromhttps://esgf-node.llnl.gov/projects/cmip6/

[7]

10.1007/0-387-34471-3

[8]

10.1016/j.gloplacha.2006.11.030

[9]

10.1007/978-0-387-74749-1

[10]

Falcon W. (2019)

[11]

10.1007/s00382-021-05869-8

[12]

10.5194/esd-11-537-2020

[13]

Assessing the impacts of 1.5 °C global warming – simulation protocol of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP2b)

Katja Frieler, Stefan Lange, Franziska Piontek et al.

Geoscientific Model Development 10.5194/gmd-10-4321-2017

[14]

GAN training code [Software]. (2023).Zenodo Repository. Retrieved fromhttps://doi.org/10.5281/zenodo.8203912 10.5281/zenodo.8203912

[15]

Goodfellow I. (2014)

[16]

10.1098/rsta.2020.0092

[17]

10.5194/hess-16-3383-2012

[18]

10.1175/1525-7541(2001)002<0406:mspoah>2.0.co;2

[19]

Harris L. McRae A. T. Chantry M. Dueben P. D. &Palmer T. N.(2022).A generative deep learning approach to stochastic downscaling of precipitation forecasts. arXiv preprint arXiv:2204.02028. 10.1029/2022ms003120

[20]

10.1002/qj.3803

[21]

Deep Learning for Improving Numerical Weather Prediction of Heavy Rainfall

Philipp Hess, Niklas Boers

Journal of Advances in Modeling Earth Systems 10.1029/2021ms002765

[22]

10.1038/s42256-022-00540-1

[23]

10.1175/1520-0477(1997)078<0005:tgpcpg>2.0.co;2

[24]

10.1038/s41598-021-85405-0

[25]

IPCC (2021)

[26]

Kingma D. P. &Ba J.(2014).Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.

[27]

10.1038/s41586-021-04283-8

[28]

Krasting J. P. (2018)

[29]

Trend-preserving bias adjustment and statistical downscaling with ISIMIP3BASD (v1.0)

Stefan Lange

Geoscientific Model Development 10.5194/gmd-12-3055-2019

[30]

Lange S.(2022).ISIMIP3BASD[Software].Zenodo.https://doi.org/10.5281/zenodo.6758997 10.5281/zenodo.6758997

[31]

Deep learning

Yann LeCun, Yoshua Bengio, Geoffrey Hinton

Nature 10.1038/nature14539

[32]

Lorenz E. N. (1996)

[33]

10.1126/science.235.4792.1036

[34]

10.1103/physreva.45.6989

[35]

The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6

Brian C. O'Neill, Claudia Tebaldi, Detlef P. van Vuuren et al.

Geoscientific Model Development 10.5194/gmd-9-3461-2016

[36]

Price I. (2022)

[37]

Data‐Driven Medium‐Range Weather Prediction With a Resnet Pretrained on Climate Simulations: A New Model for WeatherBench

Stephan Rasp, Nils Thuerey

Journal of Advances in Modeling Earth Systems 10.1029/2020ms002405

[38]

10.1038/s41586-021-03854-z

[39]

10.1109/tgrs.2010.2103946

[40]

Schneider U. "GPCC full data reanalysis version 6.0 at 0.5°: Monthly land‐surface precipitation from rain‐gauges built on GTS‐based and historic data" GPCC Data Report (2011)

[41]

The Double‐ITCZ Bias in CMIP3, CMIP5, and CMIP6 Models Based on Annual Mean Precipitation

Baijun Tian, Xinyu Dong

Geophysical Research Letters 10.1029/2020gl087232

[42]

10.1007/s00382-020-05447-4

[43]

10.1038/s41467-021-27111-z

[44]

10.5194/hess-22-3175-2018

[45]

10.1007/s00382-022-06277-2

[46]

10.5194/gmd-16-535-2023

[47]

10.1073/pnas.1312330110

[48]

10.1029/2020ms002109

[49]

WFDE5 over land merged with ERA5 over the ocean (W5E5 v2.0) [Dataset]. (2021).ISIMIP Repository. Retrieved fromhttps://doi.org/10.48364/ISIMIP.342217 10.48364/isimip.342217

[50]

10.1175/jcli3987.1

Showing 50 of 52 references

Cited By

23

Current progress in subseasonal-to-decadal prediction based on machine learning

Zixiong Shen, Qiming Sun · 2024

Applied Computing and Geosciences

Metrics

23

Citations

52

References

Details

Published: Oct 01, 2023
Vol/Issue: 11(10)
License: View

Authors

P

Philipp Hess

Earth System Modelling School of Engineering and Design Technical University of Munich Munich Germany; Potsdam Institute for Climate Impact Research Member of the Leibniz Association Potsdam Germany

S

Stefan Lange

Potsdam Institute for Climate Impact Research Member of the Leibniz Association Potsdam Germany

C

Christof Schötz

Potsdam Institute for Climate Impact Research Member of the Leibniz Association Potsdam Germany

N

Niklas Boers

Earth System Modelling School of Engineering and Design Technical University of Munich Munich Germany; Potsdam Institute for Climate Impact Research Member of the Leibniz Association Potsdam Germany; Department of Mathematics Global Systems Institute University of Exeter Exeter UK

Funding

Volkswagen Foundation

Horizon 2020 Framework Programme Award: 820970

Cite This Article

Philipp Hess, Stefan Lange, Christof Schötz, et al. (2023). Deep Learning for Bias‐Correcting CMIP6‐Class Earth System Models. Earth's Future, 11(10). https://doi.org/10.1029/2023ef004002

Deep Learning for Bias‐Correcting CMIP6‐Class Earth System Models

You May Also Like