Continuous Distributed Modeling of LID/GI: Scaling from Site to Watershed

摘要

Better insight into the benefits of LID/GI practices can be gained with a continuous distributed physics-based model that allows us to scale-up from the individual practice to watershed response within a single modeling framework. The conservation laws of rainfall-runoff are used to formulate watershed response for prior conditions and post-construction of LID/GI. Weather radar, when properly quality controlled and enhanced for accuracy, accounts for the local/regional weather affecting the natural, urbanized, and/or modified hydrology of the watershed. By using continuous rainfall data, the modeling and our understanding of how practices affect the hydrologic response is extended beyond single frequency-based approaches. Commonly used synthetic hyetographs, such as the SCS Type Ⅰ, Ⅱ, or Ⅲ temporal distributions, are based on the 25-yr storm frequency depths. With actual rainfall time-series, understanding can be gained as to performance given realistic mean inter-event arrival, depth, duration, and intensity from the historic record. Continuous simulation for a typical year or multiple years reveals how these practices will perform under realistic inputs and with physics-based model assumptions. When sizing LID/GI practices and infrastructure, and in particular, detention volume using synthetic hyetographs as model input rather than using continuous actual rainstorms can result in oversizing of outlet works, and reduced settling times, while underestimation of storage volume can result if successive storms are not taken into account.