Modeling Direct Runoff Hydrographs with the Surge Function

Voytenko, Denis

Bve-upwardtuent, individual peaks. T the unknown parameters for each veriduals. The results allow goodneas a comparison of peaks at the ds to ve rivers from three distinct hydrologic f the gamma distribution, which is known to be u to 15-minute hydrograph data.ASTRACT: A surge function is a mathematical function of the form f(x) = axpe-bx. We simplify the surge function by holding p constant at 1 and investigate the simplified form as a potential model to represent the full peak of a stream discharge hydrograph. The previously studied Weibull and gamma distributions are included for comparison. We develop an analysis algorithm which produces the best-t parameters for every peak for each model function, and we process the data with a MATLAB script that uses spectral analysis to filter year-long, 15-minute, stream-discharge data sets. The filtering is necessary to locate the conca inflection points used to separate the data set into its constihe Levenberg-Marquardt algorithm is used to iteratively estimatesion of the modeled peak by minimizing the sum of squares of resss-of-t comparisons between the three model functions, as well same gage through the year of record. Application of these methoregions shows that the simple surge function is a special case oseful as a modeling function for a full-peak hydrograph. The study also confirms that the Weibull distribution produces good ts