Hydro-informatics and Innovative Technology
Green Infrastructures (GI) has been widely used to solve urban nonpoint source pollution problems in the U.S. In practice, GI planning faces two challenges: to achieve load reduction targets with least cost through spatially arranging different types and sizes of GI projects, and to achieve optimality in time through prioritizing GI projects implementation order. Although tools and methods have been developed during the past decade for optimal GI planning, the success of these tools, such as the U.S. EPA’s SUSTAIN, is limited to spatial optimization at relatively small scales or coarse resolution. No study has been reported to address large scale spatial simulation-optimization of GI, let alone achieving the optimality in both space and time due to the computational complexity involved in such problems.
In the Upper Chollas watershed in San Diego, 7664 potential GI project sites were identified based on high-resolution Lidar data and areal map, and the task of the GI planning is to identify the cost-effective GI arrangement that can achieve the 29% Zinc load reduction from the watershed. This problem is equivalent to a simulation-optimization model consisting of more than 7000 decision variables, which is far beyond the solving capability of the widely applied SUSTAIN or its peers. To deal with the large scale GI optimization problem, we developed a multiple tiered optimization approach to overcome the computational bottleneck in solving such a problem. The approach involves progressively formulating the GIs at atomic level, molecular level and then the integrated level, and the optimal solutions were sequentially obtained to achieve the load reduction target with least cost. Upon the completion of the spatial optimization, a series of temporal optimization models were then formulated to prioritize the implementation order of the GI projects for cost-effectiveness over time. The study identified an optimal arrangement of GIs which reaches the 29% load reduction target with only about 20% of the pre-optimization cost, saving hundreds millions dollars. The optimal GI implementation plan was mapped to individual GI projects and organized in a web-based decision support platform to support the city’s real time decision and scheduling on construction of individual GI project.
Wednesday, January 4
1:30 PM – 3:00 PM
Thursday, January 5
10:00 AM – 5:00 PM