Deep learning with satellite images enables high-resolution income estimation: A case study of Buenos Aires

Nicolás F Abbate, Leonardo Gasparini, Franco Ronchetti and Facundo M Quiroga

PLOS ONE, 2026, vol. 21, issue 1, 1-34

Abstract: High-resolution income data is crucial for informing policy decisions as it allows policymakers to better understand the distribution of wealth and poverty. However, obtaining this information is often cost-prohibitive, especially in developing countries. We evaluate the potential of using high-resolution satellite imagery and machine learning techniques to create income maps with a high level of geographic detail. We train a neural network with satellite images from the Metropolitan Area of Buenos Aires (Argentina) and 2010 census data to estimate per capita income at a 50x50 meter resolution for 2013, 2018 and 2022. The model, based on the EfficientNetV2 architecture, demonstrates strong predictive accuracy for household incomes (R2 = 0.878), achieving a spatial resolution over 20 times finer than existing methods in the literature. The model also allows estimating income maps for arbitrary images, and can therefore be applied at any point in time. Our approach opens up new possibilities for generating highly detailed data, which can be used to assess public policies at a local level, target social programs more effectively, and address information gaps in areas where traditional data collection methods are lacking.

Date: 2026
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0338110 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 38110&type=printable (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:plo:pone00:0338110

DOI: 10.1371/journal.pone.0338110

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().