Building Knowledge Discovery and Information Fusion Tools for Collaborative Systems to Adaptively Manage Uncertain Hydrological Resources

点击翻译按钮获取中文摘要

There is a critical need to accurately and efficiently assess and manage water quantity. This challenge is made greater because water management must be conducted under conditions of uncertainty about current and future water resources. Adaptive water management has become the policy heuristic adopted for flexible water management that responds to ever changing physical and social demands. The central challenge of the Adaptive Management approach is the need to quantify the uncertainty in observed water measurements. This research will create new knowledge discovery and information fusion algorithms to compute two watermetrics and embed them in an overall knowledge management system. The long-term goal is to develop intelligent, scalable decision support tools for collaborative systems that explicitly incorporate uncertainty to analyze and integrate hydrological data and information. The specific objectives of the proposal are:1. Knowledge Discovery and Information Fusion: Design scalable algorithms to (a) compute Quality of Data (QoD) Quality of Information (QoI) and Quality of Knowledge (QoK), (b) incorporate proximity in geospatial events into spatio-temporal data mining and (c) integrate heterogeneous, incomplete, and uncertain geospatial data and information.2. Water Metrics: Develop an integrated water metric on Water Status - the integrated state of all water resources at a given point in space and time.3. Adaptive Management: Identify specific Water Metrics that managers and decision makers use for adaptive water management.4. Software Products: Build a suite of web-based knowledge discovery and information fusion tools to analyze, integrate and visualize hydrological data, water metrics and their quality. Intellectual Merit: The proposed research will contribute knowledge discovery and information fusion methods that explicitly model and propagate uncertainty in general and geospatial analysis in particular. An integrated framework of a digital repository of multiple and diverse data sets will be designed, powered by effective and efficient knowledge discovery, information fusion and visualization tools, that: i) assists hydrologists develop comprehensive models and metrics for the analysis of the water cycle; ii) supports decision makers in adaptive management decisions; and iii) helps policy makers study the social and legal impact on water users.Broader Impact: The research will advance the state of the art in three disciplines: (a) computer science, by developing novel knowledge discovery techniques, (b) hydrology, by developing two integrated water metrics using information fusion approach, and (c) water management and other applications, by identifying important parameters for successful adaptive management.