GEOGRAPHICAL INFORMATION SYSTEM BASED HYDROLOGICAL MODELLING OF CLIMATE CHANGE VARIABLES ON HADEJIA RIVER SUB-BASIN, NIGERIA

GEOGRAPHICAL INFORMATION SYSTEM BASED HYDROLOGICAL MODELLING OF CLIMATE CHANGE VARIABLES ON HADEJIA RIVER SUB-BASIN, NIGERIA

ABSTRACT

Soil and Water Assessment Tool (SWAT) and Geographical Information System (GIS) based hydrological modelling was applied to Hadejia River Sub-basin- an integral of the Chad Basin in the Northern Nigeria- to determine the impacts of land use/cover dynamics on the hydrological regime of the sub-basin under the changing climate factors. The study enhanced solution with dx understanding to the complex problems in the sub-basin, with large quantities of data associated with water systems and distributed nature of hydrological elements. Field data of soil sample, stream flow record from1978 to 2014, rainfall and temperature field data from 1979 to 2015 obtained from NiMET, Spatial and temporal satellite radar hydrometeorology data from 1979 to 2013, satellite imagery of land use/cover digital map for the year 1975, 2000 and 2013, digital soil map of 2013, Digital Elevation Model (DEM) of 30m resolution of the sub-basin were obtained from waterbase.org and used in the modelling. The daily radar weather data of rainfall and temperature was adjusted using the field data of rainfall and temperature, then applied in the modelling. GIS software was used to process, project and characterize the sub-basin physiographical data (Digital Elevation Model-DEM, soil data, and land use/cover and stream network) into WGS 1984 UTM Zone 32N. The point (outlet point) where the field stream flow data was obtained along the river was used to automatically delineate the watershed extent. SWAT was used to simulate and estimate the basin hydrologic parameters that include water yield, ground water recharge, lateral flow, surface runoff, and sediment yield, losses due to potential evapotranspiration and evaporation under the three different land use/cover maps of 1975, 2000 and 2013. In R programming, the trend in the field data of rainfall and temperature time series was determined using the Mann Kendall, Sen.slope and Pettitt change point nonparametric test, as recommended by World Metrologic Organization. The predictive model for the sub-basin field temperature and precipitation time series data was determined by the application of Seasonal Auto Regressive Integrated Moving Average Model (Seasonal ARIMA Model), the data was divided into training dataset (1982 to 2010) for model estimation and testing dataset (2011 to 2014) for model validation. By the application of SWAT Calibration and Uncertainty Procedures (SWATCUP), the observed stream flow data sensitivity parameters were determined, calibrated and validated. Correlation coefficient of one (1) was obtained for the ix adjusted radar data of rainfall and temperature with the respective field data. The delineated study area extent was estimated at 15,663.02 km2 ; perimeter at 1,323.51km; spot height from 784m to 351m; The longest river (Hadejia River) has total length of 364.461km, a 3rd order stream; principally gentle slope from 0.20-0.92. Based on USDA soil taxonomy, the soil imagery and validated field soil data, the study area has three dominant soils; “sandsandyloam”, “sandyloam” and “sandyloam-sandyclayloam”. The characteristics of the land use/cover dynamics from 1975 to 2013 with respect to the area occupied in percentage are; Range Land -60.74, Riparian Forest -4.64%, Wetland -13.58%, Forest Area -4.01%, Bare Area -16.91%, Agricultural Land +8.92%, Urban Area +24.63%, and Water Bodies +79.22% due to more dams construction. Applying the land use/cover data of 1975, 2000 and 2013 respectively, the maximum estimated mean monthly hydrological parameters per year from 1982-2013 are; Potential evapotranspiration (343.9, 343.9, 343.9), evaporation losses (117.6, 122.8, 122.7), Surface runoff (399.0, 405.8, 405.8), Ground water recharge (173.6, 169.4, 169.0), Water yield (482.4, 486.1, 486.1), Sediment yield (88.8, 83.6, 46.7) tonnes/hectare and Lateral flow (0.5, 0.5, 0.3). Using the minimum AIC values, the best ARIMA Model obtained for the seasonal prediction of the sub-basin temperature was ARIMA (3,1,3)(2,0,1)[12] and ARIMA(1,1,1)(1,1,2)[12] for rainfall. Ten (10) parameters were selected for sensitivity analysis: EPCO (plant uptake compensation factor), ESCO (soil evaporation compensation factor), CN2 (moisture condition II curve number), ALPHA_BF (baseflow alpha factor), GW_DELAY (Delay time for aquifer recharge), GWQMN (threshold water level in shallow aquifer for base flow), SPCON (Channel sediment routing parameter), SURLAG (Surface runoff lag coefficient), SEDPST_CONC (sediment deposit concetration), SOL_Z() (Depth from soil surface to bottom layer (mm)). The sensitivity analysis was followed by observed stream flow calibration and validation statistics determination. The sub-basin land use/cover dynamics shows no significant impacts on the hydrology of the Hadejia River System when compared. There was increase in potential evapotranspiration, decreases in evaporation, surface runoff, ground water recharge, basin water yield, sedimentation and lateral flow under the climate variables. High mean temperature was obtained in the months of March to May, while July to September has low records. Increases temperature was noticed beyond 1998 at the rate of 0.058. Mean monthly rainfall shows that the month of August has high rainfall while November to x March are dry months. The nonparametric test shows fluctuating increases at the rate 0.053 in rainfall that is not significant. The Seasonal ARIMA predictive model overestimate the temperature given by the mean absolute error (MAE)=1.261units, mean error (ME) of - 0.559 and root mean square error (RMSE) of 1.501. The filed rainfall was under estimated with MAE=5.458, ME=10.179 and RMSE=14.570. For 500 iterations to obtain the best simulation for observed stream flow data, the most sensitivity parameters are SURLAG (Surface runoff lag coefficient) and CN2 (moisture condition II curve number). The statistics for the SWAT Model performances based on calibration and validation respectively are: p-factor (0.76, 0.28), r-factor (0.71, 1.27), R2 (0.95, 0.62), NS (0.95, 0.32), Mean_sim(Mean_obs) 1.31(1.27) and StdDev_sim(StdDev_obs) 1.16(1.21) for calibration and Mean_sim(obs) 1.25(0.90), StdDev_sim(StdDev_obs) 1.21(1.00) for Validation. The value of R2=0.95 and NS=0.95 in the calibration stage shows a good performances of the SWAT Model, which indicates a very high goodness of fitness between the observed and stimulated stream flow. The study has demonstrated that GIS/SWAT based hydrological modeling of a watershed, can play a very significant role in basin management, by abetting watershed stakeholders to take informed rational decisions with respect to a better watershed management and planning under the impacts of land use/cover dynamics and climate variable.

BUY THIS TOPIC ON WHATSAPP

Related Project Materials

Background To The Study

With the invention of Information and Communication Technology, libraries now u...

OBJECTIVES OF THE STUDY

To develop an expert system on car maintenance and troubleshooting that is capable of assisting...

ABSTRACT

This study examined the impact of motivation on organizational productivity. A case study of Nigeria Bottling C...

Background of the Study

One function of the mass media which is becoming increasingly important in the modern world is the economic funct...

ABSTRACT

It has been clearly identified that one of the agent that easily deteriorate building stru...