Identifying and Categorizing Bias in AI/ML for Earth Sciences

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

doi:10.1175/BAMS-D-23-0196.1

Identifying and Categorizing Bias in AI/ML for Earth Sciences

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

Research output: Contribution to journal › Article › peer-review

10 Scopus citations

Abstract

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.

Original language	English
Pages (from-to)	E567-E583
Journal	Bulletin of the American Meteorological Society
Volume	105
Issue number	3
DOIs	https://doi.org/10.1175/BAMS-D-23-0196.1
State	Published - Mar 2024
Externally published	Yes

Keywords

Artificial intelligence
Atmosphere
Ocean
Other artificial intelligence/machine learning

Access to Document

10.1175/BAMS-D-23-0196.1

Cite this

@article{0d994ff21a8041ba8c04e6d3582a2997,

title = "Identifying and Categorizing Bias in AI/ML for Earth Sciences",

abstract = "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.",

keywords = "Artificial intelligence, Atmosphere, Ocean, Other artificial intelligence/machine learning",

author = "Amy McGovern and Ann Bostrom and Marie McGraw and Chase, \{Randy J.\} and Gagne, \{David John\} and Imme Ebert-Uphoff and Musgrave, \{Kate D.\} and Andrea Schumacher",

note = "Publisher Copyright: {\textcopyright} 2024 American Meteorological Society.",

year = "2024",

month = mar,

doi = "10.1175/BAMS-D-23-0196.1",

language = "English",

volume = "105",

pages = "E567--E583",

journal = "Bulletin of the American Meteorological Society",

issn = "0003-0007",

publisher = "American Meteorological Society",

number = "3",

}

TY - JOUR

T1 - Identifying and Categorizing Bias in AI/ML for Earth Sciences

AU - McGovern, Amy

AU - Bostrom, Ann

AU - McGraw, Marie

AU - Chase, Randy J.

AU - Gagne, David John

AU - Ebert-Uphoff, Imme

AU - Musgrave, Kate D.

AU - Schumacher, Andrea

PY - 2024/3

Y1 - 2024/3

N2 - 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.

AB - 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.

KW - Artificial intelligence

KW - Atmosphere

KW - Ocean

KW - Other artificial intelligence/machine learning

UR - https://www.scopus.com/pages/publications/85188288641

U2 - 10.1175/BAMS-D-23-0196.1

DO - 10.1175/BAMS-D-23-0196.1

M3 - Article

AN - SCOPUS:85188288641

SN - 0003-0007

VL - 105

SP - E567-E583

JO - Bulletin of the American Meteorological Society

JF - Bulletin of the American Meteorological Society

IS - 3

ER -

Identifying and Categorizing Bias in AI/ML for Earth Sciences

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this