Modular hydrologic ensemble prediction system

土地覆盖的持续变化，如城市化、工业发展、采矿等，极大地影响了加拿大河流流域的地表径流。此外，现在公认的是，这些环境变化加上气候变化正在导致地表水(河流、湖泊和水库)在地方和区域范围内发生前所未有的变化。因此，传统的水文(如径流)预报工具不能考虑流域的这种动态变化。因此，迫切需要新的能够考虑人-气候-水相互作用影响的自适应水文预报系统来提供准确可靠的水文预报，而水文预报是水资源管理的基本信息。准确可靠的水文预报对于家庭和工业供水规划、洪水预警、抗旱减灾、内河航运和水力发电至关重要。 拟议的研究计划将利用陆面监测技术(卫星、地面雷达和监测网络)的最新进展，以及气象学的进展和新兴的信息处理技术(即“顺序数据同化”和“贝叶斯预报技术”)来开发新的自适应水文预报工具。这项具有挑战性的任务将首先研究适用于分水岭尺度水文预报的稳健的顺序数据同化方法。然后，将确定的最优序贯数据同化方法与贝叶斯预报技术和物理流域模型相结合，开发出自适应水文预报系统。一个能够自动吸收各种信息源的灵活预报系统将是一个强有力的工具，可以充分利用现代陆地观测技术来考虑分水岭的动态变化，并提供更准确和可靠的水文预报。这将大大有助于加拿大更好地规划和管理水资源。

Continued changes in land cover such as urbanization, industrial development, mining, etc drastically affect the surface runoff in Canadian river basins. In addition, it is now well established that those environmental changes coupled with climate change are resulting in unprecedented changes in surface water (river flows, lakes, and reservoirs) at local and regional scales. Consequently, traditional hydrologic (e.g. streamflow) forecasting tools are not able to account for such dynamic changes in the watershed. Therefore, novel adaptive hydrologic prediction systems that can account for the impacts of human-climate-water interactions are critically needed to provide accurate and reliable hydrologic forecasts which is the essential information for water resources management. Accurate and reliable hydrologic forecasts are essential to domestic and industrial water supply planning, flood warning, drought mitigation, inland navigation, and hydroelectric power generation. The proposed research program will resort to recent advances in land surface monitoring technology (satellite, ground-based radar and monitoring networks), along with advances in meteorology, and emerging information processing technologies (namely 'sequential data assimilation', and 'Bayesian forecasting techniques') for the development of novel adaptive hydrologic prediction tool. This challenging task will be achieved by first investigating robust sequential data assimilation methods suitable for hydrologic forecasting at the watershed scale. Then, the optimal sequential data assimilation method identified will be coupled with a Bayesian forecasting technique and a physical watershed model to develop an adaptive hydrologic prediction system. A flexible prediction system that can automatically assimilate various sources of information will be a robust tool for taking full advantage of modern land observation technology to account for dynamic changes in the watershed, and to provide more accurate and reliable hydrologic forecasts. This will significantly contribute to better water resources planning and management in Canada.