Streamflow regionalization using a similarity approach in ungauged basins: Application of the geo-environmental signatures in the Karkheh River Basin, Iran

Abstract

Reliable estimation of streamflow especially in ungauged basins is most important for environmental management and planning. Regionalization of streamflow using either the physical attributes or spatial proximities of linked gauged and ungauged basins are considered to be standard techniques for estimating streamflow in ungauged basins. Unfortunately, regionalization of modeling parameters based solely on physical similarities or spatial proximity between gauged and ungauged basins is not always adequate. The hydrologic similarities between basins are related to the basin's response to precipitation, which it is referred to various conditions of basins, so considering the comprehensive characteristics from basins is very important to achieve a knowledge of the hydrology of the basin. Therefore, this study aimed to consider a comprehensive set of signatures with the fuzzy clustering approach. Similar basins were determined by considering key geo-environmental signatures, including topographical/morphological characteristics, soil properties, land use types, remote sensing (RS)-based indices, and climate indices. Redundancy in the selected geo-environmental signatures was assessed using factor analysis. Parameters of lumped and semi-distributed rainfall-runoff models were calibrated in the most similar donor basin (DB) to the ungauged basin (UB) and validated by matching them to the pseudo ungauged basin (PUB). Streamflow estimation results from the pseudo ungauged basins demonstrated that the applied procedure and semi-distributed rainfall-runoff model was an efficient method to estimate streamflow in the ungauged basins.