Two decades of land cover mapping in the Río de la Plata grassland region: The MapBiomas Pampa initiative

BAEZA, SANTIAGO; BANCHERO, SANTIAGO; VERON, SANTIAGO; OYARZABAL, MARIANO; GUERRA LARA, MARIANA; BAGNATO, CAMILO; HASENACK, HEINRICH; DE ABELLEYRA, DIEGO; SCHIRMBECK, JULIANO; WEBER, ELISEU; PETEK, MARIANA; BALDI, GERMAN; RAMOS, SEBASTIÁN; VELEZ-MARTIN, EDUARDO; GALLEGO, FEDERICO; VALLEJOS, MARIA; BARBIERI, ANDREA; SARRAILHE, SOFÍA; BRUZZONE, LAURA

Remote Sensing Applications: Society and Environment

Elsevier

Lugar: Amsterdan; Año: 2022 vol. 28

2352-9385

The Río de la Plata Grasslands (RPG) region is the largest area of the temperate humid and sub-humid grasslands biome in South America and one of the largest in the world. The region is located on fertile soils, generally very suitable for agricultural development, so it is undergoing an intense land cover change process. Our knowledge of these changes remains incomplete. Most regional-scale studies have been conducted over specific periods, limited subsets of the RGP, coarse resolution and, in general, used land cover classes that are not readily compatible. In this work we described and analyzed the land cover changes in the entire RPG region for the first two decades of the 21st century, especially those related to grasslands loss. We generated annual land cover maps with 30-m resolution that discriminate between 8 categories: native woody formation, forest plantation, swampy areas and flooded grassland, grassland, farming, non-vegetated area, water and non-observed. The map series was evaluated for the years 2001 and 2018 using a completely independent dataset, selected by stratified randomized sampling. Overall accuracy was 73.5% and 77.8% for 2001 and 2018, respectively, with user and producer accuracies that varied between classes and years. In 20 years, RPG region lost, at least, 2.4 million ha of grassland (9% of the remaining grassland area in 2001). Most of these losses are concentrated in Brazil and Uruguay and are associated with new agricultural or forestry areas that increased by 5% and 100%, respectively. Our maps allow a comprehensive understanding of the transformation processes that RPG are undergoing and provide the context on which to explore a large set of hypotheses related to ecosystem structure and functioning. It will also contribute to improving decision-making at both the regional and national levels.Land use change; Landsat; Time series; Grasslands; Classification