A Potential MODIS Real-Time Product: The Vegetation Canopy Water Content
Susan
L
Ustin, UCDavis, slustin@ucdavis.edu
(Presenting)
David
Riano, UCDavis, driano@ucdavis.edu
Marco
Trombetti, UCDavis, mtrombetti@cstars.ucdavis.edu
Vegetation water stress drives wildfire behavior and risk, having important implications for biogeochemical cycling in natural ecosystems, agriculture, and forestry. Water stress limits plant transpiration and carbon gain. The regulation of photosynthesis creates close linkages between the carbon, water, and energy cycles and through metabolism to the nitrogen cycle. Radiative transfer models, such as PROSPECT-SAILH, determine how sunlight interacts with plant and soil materials. These models can be applied over a range of scales and ecosystem types. Artificial Neural Networks (ANN) were used to optimize the inversion of these models to generate a vegetation Canopy Water Content product (CWC) from MODIS. We carried out multi-scale validation of the product using field data, airborne and satellite cross-calibration for the USA. The CWC product inputs are 1) The MODIS Terra/Aqua surface reflectance product 2) The MODIS land cover map reclassified to grassland, shrub-land and forest canopies 3) An ANN trained with PROSPECT-SAILH 4) A calibration file for each land cover type. An implementation plan for the direct read out MODIS CWC has been recently approved. CWC estimates will help predicting linkages between biogeochemical cycles, which will enable further understanding of feedbacks to atmospheric concentrations of greenhouse gases. It will also serve to estimate primary productivity of the biosphere monitor/assess natural vegetation health related to drought, pollution or diseases improve irrigation scheduling. These estimates will also allow researchers to identify wildfire behavior/risk: drives ignition probability and burning efficiency to be used as an indicator of soil moisture and Leaf Area Index.
NASA Carbon Cycle & Ecosystems Active Awards Represented by this Poster: