Effects of rainfall variability on spatial accumulation of peak runoff and excess runoff depth: Little Washita River Basin, Oklahoma, USA


Komuscu A. Ü. , Legates D. R.

Journal of Environmental Hydrology, vol.7, 1999 (Scopus) identifier

  • Publication Type: Article / Article
  • Volume: 7
  • Publication Date: 1999
  • Journal Name: Journal of Environmental Hydrology
  • Journal Indexes: Scopus
  • Ankara Haci Bayram Veli University Affiliated: Yes

Abstract

This paper demonstrates the spatial variability of surface runoff components, accumulated peak flow (Qa) and excess runoff depth (Ra), in response to varying distribution of precipitation on a basin scale. A hydrologic representation of the Little Washita Basin, Oklahoma, USA, was developed using HEC-1, a lumped parameter-based single event model. Accumulation of peak flow and excess runoff was computed at 20 locations along the drainage network. The analysis focused on the differences in the spatial and temporal distribution of precipitation while total basin rainfall and basin hydrologic conditions are held constant. The study also relates variability in the surface flow to the storm's duration and depth. The analysis shows that heterogeneous rainfall intensities in both space and time greatly influence peak flow. Uneven spatial distribution of precipitation directly contributes to higher peak flows, particularly in storms of short duration. The larger variabilities with Qa are observed when rainfall has a nonuniform distribution and high intensities. When the distribution of rainfall was more even, the resulting Qa and Ra showed less variability. The contribution from new subwatersheds to Qa in a downstream direction is usually less if the high intensity rainfall areas are located far from the centroid of the basin. Coupled with the steady increase in drainage area in the downstream direction, changes in the magnitude and variability of Qa decrease. Proximity of high intensity rain cells to basin outlet also gains major importance in the spatial behavior of Ra. The study also shows that for high storm totals, flow volumes and peak flow can be simulated more accurately compared to low storm totals.