HYDROLOGICAL MODELING FOR FLOOD FLOWS

Abstract

From ancient times, floods have caused colossal damages to the life and property throughout the world. In the tropical countries people had suffered a great deal on accounts of floods. Problem has become much more sever with the explosion in population. To meet the ever-increasing demand huge water storage structures on river system have been constructed. This has resulted into heavy reductions in the flows downstream of rivers. As a consequence, people started encroaching the flood plain which in fact is the right of the river to spill its excess water beyond the river valley i.e. overand above the natural levees. However, when large amounts of the regulated flows are allowed to pass through the regulated water storage structures, the settlements in the floodplains suffer the most. This study aims at developing a rainfall-runoff-flood inundation model for ungauged or partially gauged catchments. The objectives of this study have been clearly specified in Chapter 1. The technical tools pertaining to model development along with the software used have been summarized in Chapter 2, which describes the architecture of methodologies. A chapter completely devoted towards review of the literature has been dispensed with keeping in view the diversified nature of the technical aspects involved. Instead, detailed literature review is given in the beginning of the chapters on various aspects of the thesis. The proposed model has been applied onto the upper Ganga basin of the Ganga river in India. Little data is available for this catchment as it has been categorized as "Classified and Secret". Therefore, the catchment is treated as ungauged or partially gauged. The salient physiographic features ofthis catchment have been derived from freely available data (viz. Shuttle Radar Topography Mission data (SRTM) and Enhanced Thematic Mapper Plus (ETM+) Satellite Imagery). The same are given in Chapter 3. A comprehensive analysis of input rainfall pertaining to Intensity Duration Frequency (IDF) analysis has been reported in Chapter 4. These calculations have been carried out using L-moment approach. In majority cases, Pearson Type III (PT3) distribution fitted best and used for extrapolation of IDF curves prepared for the upper Ganga basin. For the first time in the surface hydrology, transformation of rainfall into runoff has been attempted by developing a Geomorphologic Instantaneous Unit Hydrograph (GIUH) model using the two-parameter Nakagami-w distribution. Details of the methodology involved for the ungauged catchments have been discussed in Chapter 5. The sensitivity analysis reveals that among different geomorphologic parameters, length of the highest order stream, Ln, happens to be the most sensitive parameter of this model. Moreover, for gauged catchments, the Nakagami-/w distribution optimized by Genetic Algorithm (GA) gave satisfactory results when compared with other statistical distributions viz. two-parameter Gamma (GM), three-parameter Generalised Logistic (GLG), three-parameter Pearson Type III (PT3) and three-parameter Generalised Extreme Value (GEV) distribution. Details of the mathematical formulation of this approach are given in Chapter 6. For design flood computations, a 100-year return period rainfall of 1 hr duration is adopted. The SCS-CN method is used for rainfall excess computations. The Kinematic wave (KW) equations have been solved through Preissmann Implicit method. The most sensitive KW parameters (i.e. overland roughness r\n and channel roughness t]c) are estimated for a river stretch on Bhagirathi river which is a tributary of river Ganga. In the absence of any hydrograph, the GIUH at the upper end is considered as input, whereas, the GIUH at downstream point was taken as the output. Both these GIUHs were developed by using the Nakagami-/;? distribution. Validation of the adopted parameters is satisfactorily carried out on the sub-catchment of Alaknanda river which is another main tributary of Ganga river. The physiographic configuration of KW application for arriving at the design flood peak at the outlet i.e. at Haridwar is given in Chapter 7. Computations for the overtopping water above the levee are carried out considering it (levee) to be a broad crested weir. Conclusions and scope for future work are given in Chapter 8 which is the last chapter of this thesis.