94% of researchers rate our articles as excellent or good
Learn more about the work of our research integrity team to safeguard the quality of each article we publish.
Find out more
ORIGINAL RESEARCH article
Front. Soil Sci.
Sec. Pedometrics
Volume 5 - 2025 | doi: 10.3389/fsoil.2025.1557566
This article is part of the Research Topic Advancing spatial prediction of soil properties using remotely sensed data and geospatial artificial intelligence (GeoAI): Challenges, opportunities, and future directions View all articles
AutoRA: An Innovative Algorithm for Automatic Delineation of Reference Areas in Support of Smart Soil Sampling and Digital Soil Twins
Provisionally accepted- 1 University of Florida, Gainesville, United States
- 2 Federal Rural University of Rio de Janeiro, Seropédica, Rio de Janeiro, Brazil
- 3 Embrapa Solos, Rio de Janeiro, Rio de Janeiro, Brazil
Digital Soil Mapping (DSM) enhances the delivery of soil information but typically requires costly and extensive field data to develop accurate soil prediction models. The Reference Area (RA) approach can reduce soil sampling intensity; however, its subjective delineation may compromise model accuracy when predicting soil properties. In this study, we introduce the autoRA algorithm, an innovative automated soil sampling design method that utilizes Gower's Dissimilarity Index to delineate RAs automatically. This approach preserves environmental variability while retaining accuracy compared to an exhaustive predictive model (EPM) based on extensive sampling of the entire area of interest. Our objective was to evaluate the sensitivity and efficiency of autoRA by varying target areas (10-50% of the total area) and block size spatial resolutions (5-150 pixels) in regions of Florida, USA, and Rio de Janeiro, Brazil. We modeled a hypothetical soil property derived from a combination of commonly used DSM covariates and user inputs into autoRA. Model performance was assessed using R², root mean square error (RMSE), and Bias, aggregated into a Euclidean Distance (ED) metric. Among all configurations, the optimal RA selection-characterized by the lowest ED-was achieved with a target area of 50% and a block size of 10 pixels, closely matching the accuracy of the EPM. For example, in Rio de Janeiro, the EPM produced an ED of 0.17, while the best RA configuration yielded an ED of 0.15. In Florida, the EPM had an ED of 0.35 compared to 0.38 for the optimal RA. Additionally, the 50%-RA with a block size of 10 significantly reduced total costs by approximately 61% in Rio
Keywords: Sampling strategies, autoRA, Reference area, Digital soil mapping (DSM), smart soil sampling
Received: 08 Jan 2025; Accepted: 17 Mar 2025.
Copyright: © 2025 Rodrigues, Ceddia, Vasques, Grunwald and Babaeian. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence:
Hugo Rodrigues, University of Florida, Gainesville, United States
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.