Development of Open-book Watershed Modeling for Satellite Based Flood Forecasting in International River Basin

A parsimonious way to understand the surface hydrology across the political boundaries of a river basin is to adopt an open book watershed modeling approach, first formulated by Yen and Chow (1961). The proposed NASA mission, Global Precipitation Measurement (GPM) may now usher a new era of application of the open- book modeling framework to understand the worth of anticipated availability of high resolution satellite rainfall data for predicting transboundary river flow. In our study we developed, verified and implemented our open-book watershed model for rapid prototyping of satellite rainfall based flood monitoring systems for International River Basins (IRBs). The model follows conservation of mass and momentum balance that brings minimum requirement of data. We first demonstrate the physical consistency of our model through sensitivity analysis of some geo physical basin parameters pertinent to rainfall-runoff transformation. Next, we simulate the stream- flow hydrograph for a 4 month long period using basin-wide radar (WSR-88D) rainfall data over Oklahoma assuming an open-book configuration. Finally, using the radar-simulated hydrograph as the benchmark, and assuming a two-nation hypothetical IRB over Oklahoma, we explored the impact of assimilating NASA's real- time satellite rainfall data (IR-3B41RT) over the upstream nation on the flow monitoring accuracy of the downstream nation. We developed a relationship defining the improvement in flow monitoring that can be expected from assimilating IR-3B41RT over transboundary regions as a function of the relative area occupied by the downstream nation. The relative improvement in flow monitoring accuracy for the downstream nation can be a maximum of 45% when more than 90% of the basin is transboundary to the nation. However, flow monitoring accuracy does not appear amenable for tangible improvement when 25% or less of the basin area is transboundary to the downstream nation. For IR-3B41RT to have non-negligible improvement (> 30% reduction in Relative RMSE of streamflow prediction), the downstream nation should not occupy more than 40% of the total IRB area. Finally, we mapped this relationship world-wide on the basis of climate-regime similarity using the Koppen classification. Our climate-based mapping scheme identified 5 specific downstream nations (North Korea, Bangladesh, Senegal, Mozambique and Uruguay) that could potentially benefit significantly from the integrating basin-wide IR-3B41RT data in their flood monitoring systems.