Confirming dry-season green-up in Central Amazon forests with Landsat 8 and the role of leaf demography in Modis-MAIAC seasonal spectral patterns

Vol 19, 2019 - 96513
Oral
Favorite this paper
How to cite this paper?
Abstract

We ask: (Q1) For two Central Amazon forest regions, do Landsat 8 OLI seasonal patterns of Green chromatic coordinate (Gcc), EVI and NIR corroborate BRDF-corrected Modis-MAIAC? (Q2) What are the biological drivers of Modis-MAIAC EVI and Gcc seasonality? For Q1 we obtained, for two Landsat 8 scenes, seasonal image sets having minimal difference in their artifacts from sun-sensor geometry. We found seasonal patterns for both spectral indices and for NIR from Landsat 8 were consistent with local Modis-MAIAC, thus confirming Modis detections of dry season green-up. For Q2, we found the fraction of upper canopy crowns with young leaves (< 1mo age) was highly correlated with Modis-MAIAC Gcc. The fraction of crowns with mature leaves (2-7 mo) was highly correlated with Modis-MAIAC EVI. LAI, estimated by two methods, was poorly correlated with seasonal spectral changes.

Institutions
  • 1 Forestry / Natural resources / Michigan State University
  • 2 Instituto Nacional de Pesquisas da Amazônia
  • 3 Instituto Nacional de Pesquisas Espaciais
  • 4 Global Ecology and Remote Sensing Lab / School of Biological Sciences / The University of Hong Kong (HKU)
Track
  • Forest and other vegetation
Keywords
tropical forest green-up
EVI seasonality
phenocam
leaf demography
LAI