Hydrological modeling of ungauged wadis in arid environments using GIS: a case study of Wadi Madoneh in Jordan

  • Nezar Hammouri Department of Earth and Environmental Sciences, Hashemite University, P.O.Box 150459, Zarqa, Jordan.
  • Ali El-Naqa Department of Water Management and Environment, Hashemite University, P.O.Box 150459, Zarqa, Jordan.
Keywords: surface runoff, ungauged basin, rainfall-runoff model, digital elevation model, geographic information system, Wadi Madoneh, Jordan.


Runoff is one of the most important hydrological variables used in most of the water resources applications. Reliable prediction of runoff from land surface into streams and rivers is difficult and time consuming to obtain for ungauged basins. However, Remote Sensing (RS) and Geographic Information System (GIS) technologies can augment to a great extent the conventional methods used in rainfall-runoff studies. These techniques can be used to estimate the spatial variation of the hydrological parameters, which are useful as input to the rainfall-runoff models. The main objective of this study was to model the rainfall-runoff process in a selected ungauged basin for the purpose of groundwater artificial recharge. This model simulation was carried out using an hydrological modeling system assisted by GIS. Two model runs were carried out using precipitation data of the Intensity-Duration-Frequency (IDF) curves at Zarqa rainfall station for 10 years and 50 years return periods. With the first model run, the total direct runoff volume and the peak discharge for the 10 years return period were estimated to be 151,000 mand 5.43m3/s, respectively. For the 50 years return period, the total direct runoff volume and the peak discharge were estimated to be 280,000 m3 and 12.77m3/s, respectively. The model was optimized against observed runoff data, measured during a storm event that occurred between the 2nd and the 4th of April, 2006. The flow comparison graph indicates that the calibrated model fits well with the observed runoff data, with a peak-weighted root mean square error (RMS) of less than 2%. This calibration was performed by applying different curve numbers in the simulated model. It was possible to obtain a reasonable match between the simulated and the observed hydrographs.