Fractal Generation of Artificial Sewer Networks for Hydrologic Simulations

Abstract

Simulating urban hydrology using actual sewer networks can be tedious and even practically impossible for large and/or old areas, especially when considering high spatial resolution requirements of physically- based models. Therefore, the rapid generation of artificial networks is of considerable interest, especially for city-scale analyses. Fractals are scale-independent, self-similar geometric shapes, and fractal trees are fractals consisting of a network of connecting lines. This paper presents a new public-domain application for generating artificial sewer networks using the dendritic and space-filling 'Tokunaga' fractal tree geometry (Artificial Network Generator, ANGel). Options include varying the number of generations, clipping to a drainage basin, irregularization, and expansion (densification) of an existing network. Output consists of line, point and area shapefiles with various attributes (e.g. flow length) and a statistics (e.g. Horton's bifurcation ratio) file. An application of the program to the Faneuil Brook sub-basin in Boston is presented. Artificial sewer networks generated using the program are compared to the actual network in terms of various hydrologic statistics and travel time distributions.