ASSESSMENT OF CLIMATE CHANGE IMPACT ON RIVER FLOW REGIME AT MALAGARASI CATCHMENT-TANZANIA

Abstract

In this specific investigation to evaluate, the climate change on river flow regime a semi-distributed conceptional model was used for simulating rainfall-runoff in Malagarasi catchment-lake Tanganyika basin, Tanzania. The IPCC-DDC website was used to extract the future prediction of the climate signal of temperature and rainfall from the multi-model of the Global Climate Model (GCM) of the CMIP5. The two global climate scenarios of RCP4.5 and RCP8.5 were used to project future expectations. However, HBV Light model then forced with the ensemble mean of the downscaled daily temperature and rainfall from SDSM model to simulate the daily future runoff at the outlet of the Malagarasi catchment for near future (2020-2045), middle future (2046-2075), and late future skyline (2076-2099) of the 21st century. The downscaled rainfall have shown no change in total annual rainfall but the declining trend from January to April (during the wet season), and increasing trend of October to December and with a marginal change in May to September. Man-Kendall՚s test showed that the forecasted temperature depicts declining trend in most GCMs, insignificant at 5% significant level. The HBV (Hydrologiska Byrans Vattenbalansavdelning) light model was calibrated and validated with past rainfall, temperature and discharge data of 5 to 10 years. The satisfactory results were obtained and producing NSE between 0.85 to 0.87 for all cases. The calibration period was from 2000 to 2010 with a period of ten years and validation was from 2011 to 2015, a period of five years. Also, the coefficient of determination R2 was checked and the results were ranging between 0.86 to 0.88 for both cases. Also at the outflow of the catchment, it is noted that there is a decrease of discharge which tends to decrease yearly. The magnitude of the trend for historical data was tested using Thiel-Sen՚s Slope median estimator, which indicated that most of the trend is insignificant at 5% significant level. All scenarios of RCP4.5 and RCP8.5 from each selected GCM of CSM1.1, MIROC, BNU-ESM, and CanESM2 have shown the declining trend of river flows for the period of January to April ranging from 5.6% to 27.3% when compared with the observed period. The rest of the months shown a marginal increase in river flows ranging from 5.1% to 9.8%. The mean annual river flow depicts the declining trend running from 4.39% to 19.17% expect CSM1.1 GCM, which showed an increasing trend ranging from 9.86% to 18.47%. Furthermore, the maximum and minimum river flows during the high and low season have depicted a declining trend of 9.8% to 27.8% and 1.76% to 21.35% respectively when compared with the observed period.