Identifying and Categorizing Bias in AI/ML for Earth Sciences

McGovern, Amy; Bostrom, Ann; McGraw, Marie; Chase, Randy J.; Gagne, David John; Ebert-Uphoff, Imme; Musgrave, Kate D.; Schumacher, Andrea

doi:10.1175/bams-d-23-0196.1

i

Identifying and Categorizing Bias in AI/ML for Earth Sciences

2024
By McGovern, Amy ; Bostrom, Ann ; McGraw, Marie
Source Bulletin of the American Meteorological Society, 105(3), E567-E583

Details

Journal Title:

Bulletin of the American Meteorological Society
Personal Author:

McGovern, Amy ; Bostrom, Ann ; McGraw, Marie ; Chase, Randy J. ; Gagne, David John ; Ebert-Uphoff, Imme ; Musgrave, Kate D. ; Schumacher, Andrea
NOAA Program & Office:

CIRA (Cooperative Institute for Research in the Atmosphere)
Description:

Artificial intelligence (AI) can be used to improve performance across a wide range of Earth system prediction tasks. As with any application of AI, it is important for AI to be developed in an ethical and responsible manner to minimize bias and other effects. In this work, we extend our previous work demonstrating how AI can go wrong with weather and climate applications by presenting a categorization of bias for AI in the Earth sciences. This categorization can assist AI developers to identify potential biases that can affect their model throughout the AI development life cycle. We highlight examples from a variety of Earth system prediction tasks of each category of bias.
Source:

Bulletin of the American Meteorological Society, 105(3), E567-E583
DOI:

https://doi.org/10.1175/bams-d-23-0196.1
ISSN:

0003-0007 ; 1520-0477
Format:

pdf
Publisher:

American Meteorological Society
Document Type:

Journal Article
Rights Information:

Other
Compliance:

Library
Download URL:

https://repository.library.noaa.gov/view/noaa/69263/noaa_69263_DS1.pdf
File Type:

[PDF - 1.79 MB]
Collection(s):

NOAA Cooperative Institutes
Main Document Checksum:

urn:sha-512:906eed132847283f7d35360b7d9cb11e5618a392958fb9596a1ac5810976b79bb07b8fb204fb68e955805f994566d0c75f9fd21271f7c3c25cb4d14a46823577

ON THIS PAGE

Details

Found an issue?

Send us an email at:

noaa.repository@noaa.gov

The NOAA IR serves as an archival repository of NOAA-published products including scientific findings, journal articles, guidelines, recommendations, or other information authored or co-authored by NOAA or funded partners. As a repository, the NOAA IR retains documents in their original published format to ensure public access to scientific information.

BACK TO TOP