SPATIAL MODELING AND UNIT HYDROGRAPH DEVELOPMENT WITH RADAR RAINFALL

Abstract

Most hydrologic models use point rainfall data. Point data do not account for spatial characteristics of a storm. This research investigated the benefits of spatially- and temporally-varying rainfall data. Semivariogram analyses were made to assess the importance of the following storm characteristics: size, shape, type, and velocity. Rainfall and flow gage data from the aridlands Walnut Gulch Watershed and regional data were used. A model was developed to estimate transmission losses (TL) using hydrograph routing (temporally-varying data), then a procedure was developed to use radar rainfall data (spatially-varying data) to develop unit hydrographs (UH). Exponentially shaped UHs resulted from TLs. UHs developed from radar data agreed closely with Thiessen-averaged UHs developed from rain gage data, indicating that radar UHs better represented the overall watershed processes than a UH based on a single rain gage. Therefore, accurate UHs can be developed from radar data.