Calibration of Rainfall-Stream flow Relationship for Assessing and Forecasting Hydrological Drought in Kavir-e Lut Basin, Iran

Hydrological drought is defined as a significant decrease in the availability of water in all its forms appearing in the land phase of the hydrological cycle. These forms are reflected in various hydrological variables such as stream flow (including snowmelt and spring flow), lake and reservoir level, and groundwater level. A variety of indices and methods for characterizing hydrological drought have been devised. In this study, an index called Streamflow Drought Index (SDI) have been used for characterizing the severity of hydrological droughts. Investigation of the forecasting possibility of hydrological droughts has been carried out by two approaches. First, when the appropriate historical data are available, Markov chain method have been used. The main output of the methodology is the matrix of state transition frequency for a selected pair of reference periods under the hypothesis of a Markov chain for the underlying state process. In other words, the output is a single value of drought state while the probabilities of remaining in the same state or passing to other states in the next reference period are withdrawn from tables which have been obtained off-line. Since, in general, streamflow data are difficult to obtain in real-time, the possibility of using a meteorological drought index was investigated. More specifically, a linear function of SPI was found to predict SDI to an accuracy level which is sufficient for characterizing drought severity. This involves prior calibration of a simple regression equation with modified SPI as the explanatory variable and SDI as the explained variable. The methodology is validated using reliable data from the Nesa and Fashkoh rivers basin in the southwestern margin of Kavir-e Lut (Iran). The results indicate that calibration of rainfall and streamflow relations provides a good opportunity to forecasting of hydrological droughts states in the lack of river flow data. A key consideration is that in this basin, a high degree of successful prediction is observed for the wet period of October to March. In addition, due to lack of storage snow in this basin, which can compensate the deficit rainfall in dry period, drought states predicting for other periods is also possible by using Markov chain methodology.