Seasonal variations in phenology and productivity of a tropical dry deciduous forest from MODIS and Hyperion

Abstract

Tropical dry deciduous forests exhibit diverse phenological behaviour that has significant impact on Gross Primary Productivity (GPP). However, satellite remote sensing of GPP may be impacted by spectral and spatial resolution of sensor. The Vegetation Photosynthesis Model (VPM) was applied to time-series of Hyperion and MODIS (MOD09A1_500m) reflectance values to evaluate spatial and spectral resolution impacts on inference of seasonal variation in phenology and GPP for Teak, Bamboo and Mixed tropical deciduous vegetation ecosystems of Shoolpaneshwar Wildlife Sanctuary (SWS), Gujarat, India. Seasonal dynamics coinciding with the phenological cycle were seen in Enhanced Vegetation Index (EVI) and Land Surface Water Index (LSWI), though with greater dynamic range in Hyperion than MODIS MOD09A1_500m, as a result primarily of spectral properties. In contrast, GPP values from MOD17A2_1000m did not similarly track observed phenological changes. Hyperion EVI resampled at multiple resolutions (30m; 60m; 120m; 250m; 500m) maintained synchrony with variations in the phenological events of canopy. All three covers showed lower GPP estimates in dry season despite higher Photosynthetically Active Radiation (PAR) values suggesting that water availability rather than PAR is the critical factor governing GPP of tropical dry deciduous forests. Given the patchy landscape and moisture-driven seasonal cycle of phenology and GPP, spatial resolutions of better than 250m and narrowband spectral features like Hyperion are necessary for monitoring phenology and GPP of tropical dry deciduous forests, providing a basis for improving methodologies of photosynthesis products from upcoming EnMAP and HyspIRI programs.