| dc.description.abstract | Most hydrologic data are associated with spatiotemporal information, which is capable of presenting patterns and changes in both spatial and temporal aspects. The demands of retrieving, managing, analyzing, visualizing, and sharing these data have been continuously increasing. However, spatiotemporal hydrologic data are generally complex, which can be difficult to work with knowledge from hydrology alone. With the assistance of geographic information systems (GIS) and web-based technologies, a solution of establishing a cyberinfrastructure as the backbone to support such demands has emerged. This interdisciplinary dissertation described the advancement of traditional approaches for organizing and managing spatiotemporal hydrologic data, integrating and executing hydrologic models, analyzing and evaluating the results, and sharing the entire process. A pilot study was conducted in Chapter 2, in which a globally shared flood cyberinfrastructure was created to collect, organize, and manage flood databases that visually provide useful information to authorities and the public in real-time. The cyberinfrastructure used public cloud services provided by Google Fusion Table and crowdsourcing data collection methods to provide location-based visualization as well as statistical analysis and graphing capabilities. This study intended to engage citizen-scientists and presented an opportunity to modernize the existing paradigm used to collect, manage, analyze, and visualize water-related disasters eventually. An observationally based monthly evapotranspiration (ET) product was produced in Chapter 3, using the simple water balance equation across the conterminous United States (CONUS). The best quality ground- and satellite-based observations of the water budget components, i.e., precipitation, runoff, and water storage change were adopted, while ET is computed as the residual. A land surface model-based downscaling approach to disaggregate the monthly GRACE equivalent water thickness (EWT) data to daily, 0.125º values was developed. The derived ET was evaluated against three sets of existing ET products and showed reliable results. The new ET product and the disaggregated GRACE data could be used as a benchmark dataset for researches in hydrological and climatological changes and terrestrial water and energy cycle dynamics over the CONUS. The study in Chapter 4 developed an automated hydrological modeling framework for any non-hydrologists with internet access, who can organize hydrologic data, execute hydrologic models, and visualize results graphically and statistically for further analysis in real-time. By adopting Hadoop distributed file system (HDFS) and Apache Hive, the efficiency of data processing and query were significantly increased. Two lumped hydrologic models, lumped Coupled Routing and Excess STorage (CREST) model and HyMOD model, were integrated as a proof of concept in this web framework. Evaluation of selected basins over the CONUS were performed as a demonstration. Our vision is to simplify the processes of using hydrologic models for researchers and modelers, as well as to unlock the potential and educate the less experienced public on hydrologic models. | en_US |
