Decades of Change on the Ganges-Brahmaputra Delta; Rivers, Coastlines, Agriculture and Development

Abstract:

The Ganges-Brahmaputra delta (GBD) is formed by the convergence of two great rivers, superlative in both size and geologic activity. The GBD is home to > 150 million people with a mean population density of more than 1000 people/km². The fertile soils, abundant water and favorable climate also make the delta one of the most agriculturally diverse and productive areas on Earth. We seek to better understand the coupled natural and anthropogenic dynamics of the delta through a number of cross-disciplinary field studies of the contributing geophysical, biophysical and anthropogenic processes. To provide a synoptic multitemporal perspective for these studies, we use a variety of remotely sensed observations collected over the delta since 1988. In this overview we provide a synthesis of recent results from studies of 4 interrelated processes; river channel migration, coastal erosion and deposition, co-evolution of agriculture and aquaculture, and economic development. In each study we use multi-sensor spatiotemporal analysis of remotely sensed imagery spanning more than 2 decades to observe and quantify a variety of processes at multiple spatial and temporal scales. MODIS Enhanced Vegetation Index (EVI) composites provide 16 day temporal and 250m spatial resolution of the entire delta from 2000 to present. This image time series captures not only agricultural phenology over the entire delta but also the development of aquaculture on the lower delta and interannual fluvial dynamics on both the Ganges and Brahmaputra rivers. Landsat TM, ETM+ and OLI allow us to represent land cover as continuous fields of soil and sediment substrate, vegetation and dark surfaces (water and shadow) throughout the delta since 1988. Despite its irregular temporal sampling, Landsat’s 30 m pixels better resolve both natural and anthropogenic land cover units and its longer time series extends retrospective analyses back to 1988. More recent sensors like Ikonos, Quickbird, GeoEye and WorldView provide vicarious validation with meter-scale imagery. Annual composites of night-light brightness from the OLS and VIIRS sensors provide a complementary measure of urban growth and economic development back to 1992. The coevolution of these systems are depicted in a series of multitemporal animations spanning the past 2+ decades.