Knowledge-and data-driven approach to GIS modelling technique for groundwater potential mapping at the Upper Langat Basin, Malaysia

The traditional way of groundwater assessment for alluvial and fractured rock aquifers in Malaysia was found to be not so systematic and sometimes improper. In some circumstances groundwater exploration within hard rock aquifer had been carried out using wild cat methods. The study area is the Upper Langat Basin which is located in the south eastern part of Selangor State, Malaysia. The area coverage is about 492 km2 and approximately 27 km from the city center of Kuala Lumpur. In this study, nine groundwater controlling factors that affect groundwater occurrences are derived from remotely sensed imagery, available maps, and associated databases.Those factors are: lithology, slope, lineament, land use, soil, rainfall, drainage density, elevation, and geomorphology. Next the parameter layers were integrated and modeled by using a knowledge-driven Geographical Information System (GIS) technique. The generation of groundwater potential map was performed using four different types of GIS modelling techniques. The models were weighted linear combination (WLC), weighted aggregation method (WAM), weighted index overlay analysis (WIOA) and pairwise comparison of Analytical Hierarchical Process (AHP). The weightage and score for each parameter and their classes are based on the surveys of Malaysian groundwater expert opinion. The predicted groundwater potential map was classified into four distinct zones based on the classification scheme designed by Department of Minerals and Geoscience Malaysia (JMG). The analysis and comparison of these results show that: (1) the weightage values for groundwater controlling factors based on groundwater expert opinion indicated that the lineament length density is the most important controlling factor, followed by lithology, geomorphology, rainfall, slope, elevation, soil, drainage density, and land use, (2) the WLC classification method is considered as the best result for the knowledge-driven GIS approach with prediction accuracy of 76%, (3) the frequency ratio approach indicated that slopes between 0 and 5 degree, river density less than 0.0010 km/km2, geomorphology of floodplain and land use type of rubber have the high probability of groundwater occurrences, and (4) The WLC model also had a slightly better prediction accuracy compared to the probability approach of frequency ratio (75%) and a much better prediction accuracy compare to the existing groundwater potential map which have only 25% accurate precision. Therefore the map based on the WLC method is selected as the final map of this study. The final groundwater potential map of the study area showed that about 85.11 km2 (14.78%) of the study area falls under low potential zone, with 327.13 km2 (67.68%) on moderate potential zone, 78.37 km2 (16.90%) with high potential zone, and only 2.16 km2 (0.62%) falls under very high potential zone. The high and very high groundwater potential zones are characterized by: (1) Slope degree ranging from 0 to 5, (2) Low density of drainage systems, (3) High density of lineament, and (4) The geomorphology unit is floodplain. In addition statistical analysis indicated that the hard rock aquifer dominant of the study area is better controlled by fracture system rather than well depth. Proximity analysis of the groundwater borehole wells indicated that higher yields are located near to the lineament. Further evaluation revealed a low correlation (correlation coefficient = -0.36 and R2 = 0.13) between well yield and well depth. However there is high correlations (correlation coefficient = 0.755) between area percentage for groundwater potential zones of the final map and groundwater well yield. Results obtained from this study can be use for future planning of groundwater exploration, planning and development by related agencies in Malaysia which provide a rapid method and reduce cost as well as less time consuming. The modeling technique may also be used in other areas of similar hydrological characteristics and climate conditions.

Preview PDF FPAS 2013 9R.pdf Download (1MB) | Preview

Actions (login required)

View Item