Regional Parameterization and Applicability of Ifas Model for Flood Forecasting In Jhelum and Chenab River Basin

Abstract

Floods are considered one of the most disastrous environmental hazards world-wide, claiming millions of lives, causing widespread damage to life and property. Pakistan is located in a region which is exceedingly susceptible to climate change having its major rivers mutually shared with India as an upstream riparian. In the wake of climate change, planes of Pakistan suffer from frequent floods due to heavy monsoon (mid-June to end of September) rainfalls almost every year since last decade. Colossal devastation due to past 24 major flood events in Pakistan brought an aggregated financial loss of more than US$ 38.171 billion, human death toll of 12330; affecting approximately 197275 villages with over 616598 Sq.km of area inundated by the flood waters during the past 69 years. Almost all of Pakistan was affected by the 2010 mega flood in Indus River, driven by the exceptional monsoon rains commencing at the end of July. The exceptionally high flood of 2014 in the Jhelum and the Chenab River, triggered large scale destruction in the northeast Pakistani districts of Sialkot, Lahore, Narowal, Gujrat, Mandi Bahauddin, Gujranwala, Hafizabad, Sheikhupura, adversely affecting more than half a million people. Using numerical models is a non-structural solution for effective flood prediction and management. This study concentrates on the development and assessment of a hydrologic model based on the Public Works Research Institute-Distributed Hydrologic Model (PWRI-DHM) built in Integrated Flood Analysis System (IFAS) model for Jhelum and Chenab River basins with a total area of 100, 940 km2. IFAS model was applied to simulate three flood events of 2014, 2015 and 2017. A grid size of 5×5km was considered. The 30 sec global datasets of DEM and land cover based on global map provided by ISCGM were up scaled to 5×5km grid. Soil textural classes of Food and Agriculture Organization’s (FAO/UNESCO DSMW) created digital soil map of the world were input for the model. A 2-layered spatially distributed tank model built in IFAS was selected for the flood analysis. Various rainfall sources like GSMaP-NRT, and 3B42RT were corrected with ground observatory rainfall using different correction methods and tested in the model. Simulated hydrographs were compared with measured hydrographs at control stations namely Rasul Barrage, Khanki Barrage, Qadirabad Barrage and Trimmu Barrage. In addition snowmelt discharges were also computed by the model and snowmelt curves were developed. Dam and Barrages operations were integrated with rainfall-runoff analysis and snowmelt to generate hydrographs. In this study, model was calibrated for vii vii two flood events (2014 & 2015), and resulting optimized parameter were validated for recent flood year of 2017. The results showed that GSMaP-NRT satellite rainfall was insufficient in volume compared with ground observatory. GSMaP-IF rainfall correction method showed poor performance owing to the lack of availability of ground observatory rainfall data for the trans-boundary portion of the basin. Type-1 rainfall correction method showed very good results for all the five stations except for confluence point at Trimmu Barrage where complex flow conditions were not properly replicated by the model. The model succeeded in replicating all magnitude of floods using Type-1 rainfall correction method. Inclusion of dam and barrages in the model improved the simulated flow results. 3B42RT satellite rainfall underestimated the measured flows but showed good match in comparison with GSMaP-IF rainfall. This concludes that the satellite rainfall estimates need to be calibrated or corrected especially for rainfall volumes before input for model. Snowmelt module of IFAS was successfully implemented which estimated the snowmelt contribution as 3-7% and 4-23% of the average daily discharge during the monsoon season at Mangla and Marala respectively during 2014 and 2015.Nevertheless, the model has been successful in replicating the measured flood peaks highlighting its effectiveness as a tool to help manage floods.