A perspective on the interpretability of poverty maps derived from Earth Observation

Gary Watmough (), Dan Brockington, Charlotte L.J. Marcinko, Ola Hall, Rose Pritchard, Tristan Berchoux (), Lesley Gibson, Enrique Delamonica, Doreen Boyd, Reason Mlambo, Seán Ó Héir and Sohan Seth
Additional contact information
Gary Watmough: The University of Edinburgh
Dan Brockington: ICTA - Institut de Ciencia i Tecnologia Ambientals - UAB - Universitat Autònoma de Barcelona = Autonomous University of Barcelona = Universidad Autónoma de Barcelona, ICREA - Institució Catalana de Recerca i Estudis Avançats = Catalan Institution for Research and Advanced Studies
Ola Hall: Lund University
Rose Pritchard: University of Manchester [Manchester]
Tristan Berchoux: CIHEAM-IAMM - Centre International de Hautes Etudes Agronomiques Méditerranéennes - Institut Agronomique Méditerranéen de Montpellier - CIHEAM - Centre International de Hautes Études Agronomiques Méditerranéennes, UMR TETIS - Territoires, Environnement, Télédétection et Information Spatiale - Cirad - Centre de Coopération Internationale en Recherche Agronomique pour le Développement - AgroParisTech - CNRS - Centre National de la Recherche Scientifique - INRAE - Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement
Lesley Gibson: SU - Stellenbosch Business School [Stellenbosch University] - SU - Stellenbosch University - Universiteit Stellenbosch [South Africa]
Enrique Delamonica: UNICEF [New York, NY, USA]
Doreen Boyd: UON - University of Nottingham, UK
Reason Mlambo: The University of Edinburgh
Seán Ó Héir: The University of Edinburgh
Sohan Seth: The University of Edinburgh

Post-Print from HAL

Abstract: The use of Earth Observation Data and Machine Learning models to generate gridded micro-level poverty maps has increased in recent years, with several high-profile publications. producing some compelling results. Poverty alleviation remains one of the most critical global challenges. Earth Observation (EO) technologies represent a promising avenue to enhance our ability to address poverty through improved data availability. However, global poverty maps generated by these technologies tend to oversimplify the complex and nuanced nature of poverty preventing progression from proof-of-concept studies to technology that can be deployed in decision making. We provide a perspective on the EO4Poverty field with a focus on areas that need attention. To increase the awareness of what is possible with this technology and reduce the discomfort with model-based estimates, we argue that the EO4Poverty models could and should focus on explainability and operationalizability alongside accuracy and robustness. The use of raw imagery in black-box models results in predictions that appear highly accurate but that are often flawed when investigated in specific local contexts. These models will benefit for incorporating interpretable geospatial features that are directly linked to local context. The use of domain expertise from local end users could make model predictions accessible and more transferable to hard-to-reach areas with little training data.

Date: 2025-12
Note: View the original document on HAL open archive server: https://hal.science/hal-05742083v1
References: Add references at CitEc
Citations:

Published in Science of Remote Sensing, 2025, 12, pp.100298. ⟨10.1016/j.srs.2025.100298⟩

Downloads: (external link)
https://hal.science/hal-05742083v1/document (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-05742083

DOI: 10.1016/j.srs.2025.100298

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().