Assessment of climate change impact on the Meghna river basin using GBHM (geomorphology based hydrological model)

Abstract

Hydrological models, used to study the behavior of river networks, have developed significant progress over the past four decades. Among several distributed hydrological models, Geomorphology Based Hydrological Model (GBHM) is utilized in this study. Hillslope elements are the fundamental computational units of the model. The catchment is divided into a number of flow intervals. Each interval is represented by its area and width function. This scheme of topography representation has made GBHM suitable for complete hydrological simulation of large river basins. GBHM is used in this thesis to study the Surma-Meghna basin which is the longest (669 km, inside Bangladesh) river system among the four major river systems of Bangladesh and covers a large area. The system also drains out one of the world’s heaviest rainfall areas (Cherapunji, Meghalaya, India). Therefore, study on the Meghna river basin allows the understanding of the hydrological processes affecting a vast region of the country. Observed rainfall data is used for simulation for benchmark of the model. The simulated discharge is compared with the observed discharge. After verification of the model, it is used to predict future discharge of the river network. The simulated discharge hydrograph using TRMM measured rainfall data shows oscillations around the observed discharge values due to the assumption made in GBHM. The overshootings are from the over estimation of TRMM rainfall. So, a combined scheme is used to minimize the effect of limitations of TRMM data. Observed rain gauge data are used wherever available and TRMM data are used for the remaining area. Then, overshootings are greatly reduced and the ratio of absolute error to the mean introduced by World Meteorological Organization becomes 0.38 which is within the allowable error ratio limit for large basins. Finally, PRECIS predicted rainfall data is used for discharge estimation. The average daily simulated discharge for the years 2001, 2031, 2061 and 2091 are 6370, 4050, 4650 and 7870 cumec, respectively. The predicted discharge values in 2031 and 2061 is less than that of 2001 whereas for 2091 it is more than the other years resulting from the changed precipitation pattern due to climate change. The trend analysis shows a continuous increase in discharge. PRECIS prediction shows, climate change will cause intense precipitation in future affecting the basin runoff and river discharge. Increased discharge may cause greater number of flood events and large fluctuations in the discharge may cause water-stress in the coming years.