LARGE-SCALE MONITORING OF FOREST DISTURBANCES IN NORTHERN MATO GROSSO FROM 2000 – 2011 BASED ON THE CLOUD COMPUTED ΔRNBR INDEX

Vol 19, 2019 - 95610
Oral
Favoritar este trabalho
Como citar esse trabalho?
Resumo

This paper describes a novel approach of large-scale remote sensing - based monitoring of human-induced forest disturbances by selective logging and forest fires for the years 2000–2011 in Northern Mato Grosso State in the Brazilian Amazon, comprising more than 414,000 km2. A pixel-based yearly change detection approach is applied on multiple Landsat imagery, using a self-referenced Normalized Burn Ratio (∆rNBR) index through cloud computing with Google Earth Engine. Assessed within grid cells of 300 m ×300 m spatial resolution, the overall area of disturbed forest over 12 years covers 53,302 km2 (24,1%), thereof 38,255 km2 by selective logging (17,3%) and 18,711 km2 (8,4%) by forest fires, including 3,664 km2 (1.7%) in both categories. The yearly areas under selective logging and affected by forest fire range from 1,819 km2 (2009) to 6,984 km2 (2005) and from 68,0 km2 (2001) and 10,258 km2 (2007), respectively.

Instituições
  • 1 Joint Research Centre of the European Commission
  • 2 Instituto Nacional de Pesquisas Espaciais - SJC
  • 3 GFT Italia Srl.
  • 4 ARHS Developments S. A.
  • 5 Department of Forest Resources Management / Faculty of Forestry / The University of British Columbia
Eixo Temático
  • Degradação de florestas
Palavras-chave
forest disturbance
selective logging
forest fires
REDD+
Brazilian Amazon