A Scalable Open-source Web-analytic Framework to Improve Satellite -based Operational Water Management in Developing Countries
Nishan Kumar BiswasChair of the Supervisory Committee: Associate Professor Faisal Hossain Department of Civil and Environmental Engineering Satellite based observations are becoming more widespread for operational and decision making activities in water resources due to their near real-time availability as well as minimization of difficulties related to ground based hydrological measurements. In this study, two crucial hurdles to advance real-world operationalization of satellite datasets are addressed. First, a simple, easy-to-access, easy-to-build, robust and consistent open-source web portal is developed connecting complex models to facilitate satellite based observations by developing nations.Second, to enhance the predictability skill of satellite-based estimations, an innovative and dynamic web analytics-based correction system is developed to reduce the uncertainty related to indirect method of measurement by the satellites. The correction system comprised of precipitation correction and streamflow correction. In precipitation correction, web crawled precipitation from different sites of government agencies of South Asia region are used to estimate bias in satellite based precipitation. This dynamically bias adjusted precipitation is then used to simulate hydrological models to obtain streamflow. Simulated streamflow is corrected using discharge climatology and gauged discharge observations. These corrected datasets are finally shared through South Asian Surface Water Modelling System (SASWMS) web portal. On an average, these dynamic correction techniques reduced RMSE in streamflow by 80-90%. The take-home message is that with the growth of open-source tools, it is now possible for any resourceconstrained water management agency to build robust and cost-effective operational web portals based on satellite data and non-proprietary software. With dynamic correction features that can also be built from open-source tools, the well-known limitations of satellite-based operationalization can be overcome for water management agencies of developing countries.v