Dr. Adnan Rajib, ORISE Postdoctoral Fellow, US Environmental Protection Agency
With growing stress from natural and man-made factors, there are distinct changes in the spatio-temporal pattern of water availability and hydrologic extremes. Knowing how and where these changes are occurring as well as their future expectancies are imperative for sustainable water management and policy decisions. Hydrologic models are being increasingly used to support these prediction needs across large spatial scales. However, our prediction skills remain poor given the inadequate representation of watershed physical processes in hydrologic models and acute lack of data to verify them. As such, uncertainties in model predictions and how to reduce them have been the central problem in all hydrologic modeling efforts.
To minimize prediction uncertainties, integration of remotely sensed data with hydrologic models has emerged as an alluring solution. Progressing remote sensing techniques have led to a hypothesis that satellite-borne estimates of soil moisture, evapotranspiration and/or vegetation dynamics are proxies for on-the-ground empirical measurements, hence, their integration may compensate models’ deficiencies to accurately capture physical processes. Despite an immense potential, approaches to advance this big data integration are still in infancy. To fill this gap, I will introduce a two-step architecture: first, through assimilating multi-sensor estimates of water, energy and vegetation into the model, and second, through a spatially distributed watershed-level model calibration scheme beyond just using gaged data at discrete locations. I will further discuss a newly developed internet-based semi-automatic platform that can concurrently access multiple satellite data repositories and make these datasets ready for model integration – regardless of the spatial scale or the geographic location of model simulation. To what extent this data-model integration affects practical decisions will be highlighted by addressing three urgent areas: flood inundation pattern, impacts of changing land use on future water availability, and wetland hotspots to target restoration and protection. These will afford insights into the next-generation of eco-hydrology research, making remote sensing integration the new normal of watershed science and policy.
Dr. Adnan Rajib is a post-doctoral research fellow at the US Environmental Protection Agency’s Office of Research and Development. Immediately before coming to EPA in March 2017, Dr. Rajib earned Ph.D. in Civil Engineering from the Purdue University. In recognition of his scholarly achievements and excellence in research, he received two prestigious Purdue honors – the 2016 Bilsland Dissertation Fellowship and the Spring 2017 CE Outstanding Graduate Student Award.
Dr. Rajib’s specialization bridges multi-disciplinary problems, broadly focusing on (i) computational hydrology with satellite remote sensing integration and (ii) hydroinformatics of big geospatial data. At the EPA, he is developing hydrologic models to delineate wetland hotspots across major US river basins to target areas of wetland restoration and management. During his Ph.D., Dr. Rajib played a leadership role in NOAA – National Water Center’s 2016 Innovators Program, where he coordinated projects of national impact involving flood hazard and disaster management. In the Summer of 2015, he worked as a Visiting Research Scholar at the South Dakota State University to model the land and climate drivers of changing hydrologic regime in US Great Plains. Among other initiatives, Dr. Adnan Rajib is active in developing online tools for data-driven hydrology education.
