Mapping Irrigated Rice in Brazil Using Sentinel-2 Spectral–Temporal Metrics and Random Forest Algorithm

Brazil, a leading rice producer globally, faces challenges in systematically mapping its diverse rice fields due to varying cropping systems, climates, and planting calendars. Existing rice mapping methods often rely on complex techniques like deep learning or microwave imagery, posing limitations f...

Full description

Saved in:
Bibliographic Details
Published inRemote sensing (Basel, Switzerland) Vol. 16; no. 16; p. 2900
Main Authors Fernandes Filho, Alexandre S., Fonseca, Leila M. G., Bendini, Hugo do N.
Format Journal Article
LanguageEnglish
Published Basel MDPI AG 01.08.2024
Subjects
Algorithms
analysis-ready data
Classification
Cropping systems
Deep learning
Harvest
irrigation
Land use
Machine learning
Mapping
Microwave imagery
paddy rice
Rice
Rice fields
Time series
Brazil
Atlantic Forest
Rio Grande do Sul Brazil
Online AccessGet full text

Cover

More Information
Summary:Brazil, a leading rice producer globally, faces challenges in systematically mapping its diverse rice fields due to varying cropping systems, climates, and planting calendars. Existing rice mapping methods often rely on complex techniques like deep learning or microwave imagery, posing limitations for large-scale mapping. This study proposes a novel approach utilizing Sentinel-2 spectral–temporal metrics (STMs) in conjunction with a random forest classifier for rice paddy mapping. By extracting diverse STMs and training both regional and global classifiers, we validated the method across independent areas. While regional models tended to overestimate rice areas, the global model effectively reduced discrepancies between our data and the reference maps, achieving an overall classifier accuracy exceeding 80%. Despite the need for further refinement to address confusion with other crops, STM exhibits promise for national-scale rice paddy mapping in Brazil.
ISSN:2072-4292
2072-4292
DOI:10.3390/rs16162900