Spatial Model for Predicting the Cost of Constructing Hand-Dug Wells in Abeokuta City, Nigeria

Abstract

The heavy reliance on groundwater sources, the high cost and difficulties in accessing groundwater from a basement complex in Abeokuta necessitate the development of a simple prediction model. Hence, the objective of this paper was to develop a spatial model for prediction of the cost of constructing hand-dug wells in Abeokuta city, Nigeria, using geographic information systems. The static water level (SWL) of wells was measured across the city, and the digital terrain model (DTM) of the city was created in the GIS. Map algebra was applied to determine the depth and predict the overall cost of well construction. The map of the overburden removed was derived as the algebraic difference between the DTM and SWL extended down by the height of four concrete rings (OVERBURDEN = DTM – SWL+ 4 rings below the SWL). The map of the total cost of well construction in any part of the metropolis was produced by multiplying the sum of the cost of one unit of concrete ring and excavation of the depth of one concrete ring with the map of overburden. The model was validated by an empirical investigation of eight randomly selected wells. The results revealed relatively good accuracy, with correlation coefficients of 0.94, R2 values of 0.76, RMSE values of 0.27 and mean absolute percentage errors of 9.24 and 3.23 for the SWL and construction cost, respectively. The paper concludes that such spatial decision support is good for municipal water planning.

Keywords

• Spatial system; Predictive assessment; Hand-dug wells; Correlation coefficient