SA007-02
Quantifying Useful Wrongness: The Crucial Role of Model Metrics and Validation in Bringing Heliophysics Models from Creation to Application
Tuesday, 8 December 2020: 10:34
Virtual
Katherine Garcia-Sage1, Hyunju KIM Connor2, Anna DeJong3, Alexa Jean Halford4, Maria M Kuznetsova5, Michael Warren Liemohn6, M. Leila Mays5, Dogacan Su Ozturk7, Lutz Rastaetter4, Robert J Redmon8 and Robert M Robinson9, (1)Goddard Space Flight Center, Greenbelt, MD, United States, (2)University of Alaska Fairbanks, Fairbanks, AK, United States, (3)Catholic University of America, Physics, Washington, DC, United States, (4)NASA Goddard Space Flight Center, Greenbelt, MD, United States, (5)NASA/GSFC, Greenbelt, MD, United States, (6)University of Michigan, Climate and Space Sciences and Engineering, Ann Arbor, MI, United States, (7)NASA Jet Propulsion Laboratory, Pasadena, CA, United States, (8)NOAA National Centers for Environmental Information, Geophysical Science Branch - Solar and Terrestrial Physics, Boulder, CO, United States, (9)NASA Goddard, Catholic University of America, Washington DC, DC, United States
Abstract:
Before our field can use cutting-edge science to address operational needs, we must build a pathway from basic research to application that goes by way of metrics and validation. Each metric or validation result is geared to answer a question - is this model
valid for this application? Understanding the use allows us to develop targeted metrics for specified operational needs. At that point, validation, especially by an independent third-party such as the Community Coordinated Modeling Center (CCMC), can be carried out and can help to level the playing field, in order to ensure that the best model for a given task is used.
I will examine the current pathways to model application using examples from the GEM/CEDAR Conductance Challenge and COSPAR ISWAT AuroraPHILE (Auroral Precipitation and High Latitude Electrodynamics) team to highlight what resources for validation our field currently has to offer, what efforts are in progress, and what is missing. These efforts highlight the need for our field to connect research scientists with “brokers” who understand current operational needs and uses. These brokers will be able to assist with ongoing work to define metrics for model applications. Standard metrics and expectations of validation as part of every well-developed research project are crucial for our field to lower the barriers along the pathway from model development to application.
