Estimating ungauged stream flows based on model regionalization: examples from the mountainous, semi-arid Karkheh River Basin, Iran

Masih, Ilyas; Uhelnbrook, S.; Maskey, S.; Ahmad, Mobin-ud-Din; Islam, Aminul. 2008. Estimating ungauged stream flows based on model regionalization: examples from the mountainous, semi-arid Karkheh River Basin, Iran. In Brhuthans J.; Kovar, K.; Hrkal, Z. (Eds.). HydroPredict 2008 Conference on Predictions for Hydrology, Ecology, and Water Resources Management: Using Data and Models to Benefit Society, Prague, Czech Republic, 15-18 September 2008. pp. 7-10.

The study examines the possibility of simulating time series of stream flows for ungauged catchments based on hydrological similarity. As an example the mountainous, semiarid Karkheh river basin (50,764 km2) of Iran is presented. The frequently applied HBV model was applied to simulate daily stream flow with parameters transferred from gauged catchment counterparts. Hydrological similarity is defined based on three similarity measures: geographical area, spatial proximity and shape of the flow duration curve (FDC). FDCs for the ungauged catchments were predicted using logarithmic relationship derived from physiographic characteristics of eleven gauged catchments. The study shows that transferring HBV model parameters based on the FDC similarity criterion produces better runoff simulation compared to the similarity criteria based on catchment area and geographical proximity. The validation of the catchment similarity analysis using the FDC on monthly and daily flow simulation resulted in mean Nash-Sutcliffe model efficiency, Reff, of 0.70 and 0.57, respectively. The study concludes that the methods of utilizing FDCs could be applied for estimating ungauged stream flows in the mountainous parts of the Karkheh river basin and source area of other major rivers in that region (e.g. Dez, Karun and Zayandeh Rud).