Using an Explainable Machine Learning Approach to Characterize Earth System Model Errors: Application of SHAP Analysis to Modeling Lightning Flash Occurrence

Computational models of the Earth System are critical tools for modern scientific inquiry. Effortstoward evaluating and improving errors in representations of physical and chemical processes inthese large computational systems are commonly stymied by highly nonlinear and complexerror behavior. Recen...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Journal of advances in modeling earth systems Ročník 14; číslo 4
Hlavní autoři:	Silva, Sam J, Keller, Christoph A, Hardin, Joseph
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Goddard Space Flight Center Wiley Open Access 01.04.2022 John Wiley & Sons, Inc American Geophysical Union (AGU)
Témata:	Additives Artificial intelligence Classification Cybernetics, Artificial Intelligence and Robotics Earth Earth System modeling ENVIRONMENTAL SCIENCES Errors General circulation models interpretable A.I Learning behaviour Lightning Lightning flashes Machine learning Modelling Neural networks Satellites XGBoost Shap Xgboost Artificial Intelligence Lightning Shapley Additive Explanations Machine Learning Earth System Modeling Interpretable A.i
ISSN:	1942-2466, 1942-2466
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Computational models of the Earth System are critical tools for modern scientific inquiry. Effortstoward evaluating and improving errors in representations of physical and chemical processes inthese large computational systems are commonly stymied by highly nonlinear and complexerror behavior. Recent work has shown that these errors can be effectively predicted usingmodern Artificial Intelligence (A.I.) techniques. In this work, we go beyond these previousstudies to apply an interpretable A.I. technique to not only predict model errors but also movetoward understanding the underlying reasons for successful error prediction. We use XGBoostclassification trees and SHapley Additive exPlanations (SHAP) analysis to explore the errors inthe prediction of lightning occurrence in the NASA GEOS model, a widely used Earth SystemModel. This explainable error prediction system can effectively predict the model error andindicates that the errors are strongly related to convective processes and the characteristics ofthe land surface.
Bibliografie:	GSFC Goddard Space Flight Center ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 PNNL-SA-167899 Linus Pauling Distinguished Postdoctoral Fellowship AC05-76RL01830 USDOE Laboratory Directed Research and Development (LDRD) Program
ISSN:	1942-2466 1942-2466
DOI:	10.1029/2021MS002881