Seasonal and Inter-Annual MODIS Data Responses over a Dry to Wet Amazon Tropical Forest Gradient
Alfredo
R
Huete, University of Arizona, ahuete@ag.arizona.edu
(Presenting)
Kamel
Didan, University of Arizona, kamel@ag.arizona.edu
Yosio
Edemir
Shimabukuro, INPE, yosio@ltid.inpe.br
Tomoaki
Miura, University of Hawaii, tomoakim@hawaii.edu
We investigated the seasonal dynamics of the seasonally dry to perhumid tropical forests through an east – west transect along the Amazon using high temporal frequency observations from the Terra- Moderate Resolution Imaging Spectroradiometer (MODIS). Prior studies with the NOAA-AVHRR instrument have shown the limitations of extracting seasonal signatures over the Amazon forest due to the extent of cloud contamination and the saturated response of the normalized difference vegetation index (NDVI). In this study we used 4 years of 250-m MODIS enhanced vegetation index (EVI) data, filtered with quality assurance (QA) ‘metrics’ to extract the forest signal from highly cloud-contaminated data over the Amazon. The observed MODIS temporal profiles exhibited distinct seasonal trends in the seasonally dry forests, representing the combined, integrative phenology of numerous forest tree species. The MODIS EVI did not saturate and revealed a higher response in the dry season for the seasonally dry forests, a result attributed to a flush of new biologic leaf activity. This was in agreement with net ecosystem exchange results reported from flux tower measurements. On the other hand, forest converted and disturbed areas showed a distinct drying trend with lower EVI values, due to the soil drying process and relatively shallow roots of the vegetation in the converted areas. These results show that fine temporal satellite datasets can track biologic activity in the Amazon region, including biologic responses to shifts in vegetation type and disturbance. This provides an improved understanding of the spatial and temporal variations in the carbon and water cycles throughout Amazonia.