Burned Area Mapping in the Brazilian Savanna Using a One-Class Support Vector Machine Trained by Active Fires

Autor(es)

Pereira, Allan A. ^[1] ; Pereira, Jose M. C. ^[2] ; Libonati, Renata ^[3] ; Oom, Duarte ^[2] ; Setzer, Alberto W. ^[4] ; Morelli, Fabiano ^[4] ; Machado-Silva, Fausto ^[3] ; Tavares de Carvalho, Luis Marcelo ^[5]

Afiliação do(s) autor(es)

^[1] Inst Fed Ciencia & Tecnol Sul Minas Gerais, BR-37713100 Pocos De Caldas - Brazil

^[2] Univ Lisbon, Ctr Estudos Florestais, Inst Super Agron, P-1349017 Lisbon - Portugal

^[3] Univ Fed Rio de Janeiro, Dept Meteorol, BR-21941916 Rio De Janeiro - Brazil

^[4] Inst Nacl Pesquisas Espaciais, Ctr Previsao Tempo & Estudos Climat, BR-12227010 Sao Jose Dos Campos - Brazil

^[5] Univ Fed Lavras, Dept Engn Florestal, BR-37200000 Lavras - Brazil

• Tipo de documento: Artigo Científico
• Fonte: REMOTE SENSING; v. 9, n. 11 NOV 2017.
• Citações Web of Science: 10

Resumo

We used the Visible Infrared Imaging Radiometer Suite (VIIRS) active fire data (375 m spatial resolution) to automatically extract multispectral samples and train a One-Class Support Vector Machine for burned area mapping, and applied the resulting classification algorithm to 300-m spatial resolution imagery from the Project for On-Board Autonomy-Vegetation (PROBA-V). The active fire data were screened to prevent extraction of unrepresentative burned area samples and combined with surface reflectance bi-weekly composites to produce burned area maps. The procedure was applied over the Brazilian Cerrado savanna, validated with reference maps obtained from Landsat images and compared with the Collection 6 Moderate Resolution Imaging Spectrometer (MODIS) Burned Area product (MCD64A1) Results show that the algorithm developed improved the detection of small-sized scars and displayed results more similar to the reference data than MCD64A1. Unlike active fire-based region growing algorithms, the proposed approach allows for the detection and mapping of burn scars without active fires, thus eliminating a potential source of omission error. The burned area mapping approach presented here should facilitate the development of operational-automated burned area algorithms, and is very straightforward for implementation with other sensors. (AU)

• Processo FAPESP: 15/01389-4 - Sistema brasileiro Fogo-Superfície-Atmosfera (BrFLAS)
• Beneficiário: Alberto Waingort Setzer
• Modalidade de apoio: Auxílio à Pesquisa - Regular