Home Articles Exploration of Groundwater by Integrated Modeling using Remote Sensing and Geograpical Information...

Exploration of Groundwater by Integrated Modeling using Remote Sensing and Geograpical Information System in Kedah and Perlis- Pilot study and Data Analysis

Ahmad Nadzari Yahaya
Department of Geomatic Sciences,
Universiti Teknologi MARA (UiTM) Perlis Campus,02600 Arau,
Perlis, Malaysia
E-mail : nadzari @ perlis.uitm.edu.my

Ahmad Kamal Md. Issa
Department of Civil Engineering,
Universiti Teknologi MARA (UiTM) Perlis Campus,02600 Arau,
Perlis, Malaysia
E-mail : [email protected] perlis.uitm.edu.my

Kamaruzaman Wan Yusof
Associate Professor,
Department of Civil Engineering,
Universiti Teknologi MARA (UiTM) Perlis Campus,02600 Arau,
Perlis, Malaysia
E-mail : [email protected] rocketmail.com

Zakaria Mat Arof
Associate Professor,
Department of Geomatic Sciences,
Universiti Teknologi MARA (UiTM) Perlis Campus,02600 Arau,
Perlis, Malaysia
E-mail : Dr Zakaria Mat [email protected] perlis.uitm.edu.my

Groundwater forms part of the natural water cycle and is present within underground strata. Water supply for irrigation and domestic use in Perlis state, at the most northern tip of Malaysia, is mainly supplied from surface waters, with groundwater supplementing this supply for localised uses. The geology of Perlis consist of mainly alluvium soil, granite and limestone hardrock formation. This is a pilot study to investigate the use of remote sensing data from SPOT satellite image, and other ancillary data. The analysis of these data using GIS has produced different thematic maps such as landuse, geology, rivers and drainage, land administration areas or ‘mukim’, boreholes, wells, soil types, and rainfall gauges. These maps later will be used to predict charateristics of groundwater potential zones in the study area. Groundwater potential is based on the modified DRASTIC model.
Groundwater forms part of the natural water cycle and constitutes a major portion of the cycle. Groundwater is present in various types of geological formation and occurs in permeable geologic formation called aquifers which can store and transmit water.

Water supply for irrigation and domestic use in Perlis state, Malaysia is mainly supplied from surface water from the Timah Tasuh dam in Perlis state, and the Muda and Pedu dams in neighbouring Kedah state. Groundwater supplements this supply mainly for localised uses (Syed Muhammad et.al , 2000). Malaysia is water-stressed, and recent severe droughts in the last few years have caused these dams’ water levels to drop drastically (Berita Harian-26 February 2005; The Star-18 August 2005). If groundwater can be a viable alternative source of water supply for multi-purpose usage, it can be better exploited if potential areas with abundant groundwater can be identified.

Groundwater cannot be seen directly from the earth surface, so a variety of techniques can provide information on its potential occurrence. Existing methods of identifying potential groundwater areas using geophysical and geo-electrical techniques are time consuming and costly. New technologies based on using remote sensing and Geographical Information System (GIS) was used in this pilot study to identify potential groundwater areas, in mainly alluvium soil, granite and limestone hardrock formations peculiar to Perlis state and its surrounding regions. Almost all alluvial plains have a high potential of groundwater occurrence, while in hardrock areas, groundwater is greatest in the high density lineament zones. (Khairul Anam et. al , 2000)

1.1 Objective
The objective of this pilot study is to determine the criteria needed to input into a database in order to produce thematic maps , using the technique of integration of remote sensing and GIS. These thematic maps will be used to identify charateristics and zones of potential groundwater in Perlis state.

1.2 Study Area
The study area is the whole Perlis state, situated in north- western tip of the Malaysian peninsular, bounded between latitudes 5o 05′ and 6o 35′ and between longitudes 99o 35′ and 100o 50′, with neighbouring borders of Thailand in the north and Kedah in the south. Covering an area of 795 sq. km. Perlis state comprises of paddy areas in the south and the Mata Ayer Forest Reserve in the north up to the Malaysia- Thailand borders.

It has a tropical monsoon with ‘winter winds’ (northernly wind from the Siamese Bay) with temperatures between 21 -33 degrees Celsius. Its climate experiences two distinct rainy periods (April- May, and September- October) and a prolonged drought period between December- March associated with the north-east monsoon.

In this pilot study, the first criteria is the choice and collection of input data. Secondly, GIS software (ERDAS Imagine 8.6, Arc View 3.1 and MapInfo 7.0) is used as a data processing tool, on all spatial data and attribute data.GIS is used fully to do spatial analysis on input data to produce thematic layer maps. These thematic layer maps will be used in the next stage, together with other data to produce a map of potential groundwater zones in Perlis state. Table 1 below shows the schematic outline to this approach.

TABLE 1: Methodology of Study

2.1 DRASTIC model
Based upon the DRASTIC model for evaluating groundwater pollution using hydrogeological setting, developed by the National Water Well Association in conjunction with the Environmental Protection Agency (EPA) of the United States (Aller et al,1985), the modified formula for groundwater potential (GP) by Khairul Anam et al (2000) is

GP = Rf + Lt + Ld + Lu + Te + Ss + Dd + St
Where GP= Groundwater Potential
Rf= annual rainfall
Lt= lithology
Ld= lineament density
Lu= land use
Te= topography elevation
Ss= slope steepness
Dd= drainage density
St= soil type

The DRASTIC model has been successfully used in similar studies (Khairul Anam et al, 2000) to predict groundwater potential in the Langat basin using the integration of remote sensing and GIS.
Based upon this model, initial choice of main input is spatial data consisting of a multi spectrum SPOT (Systeme Pour l’Observation de la Terre) satellite image of Perlis state (dated 2005) with a resolution of 20 meters, as in Figure 1. In a similar study to map landuse cover of northern Perlis, Kamaruzaman et al (2005) successfully utilised a Landsat TM image of Perlis state.

FIGURE 1: SPOT satellite image and Topography Map of Perlis state

Other spatial data were obtained from various government agencies as in Table 2, consisting of topography map, geology map, soil series map, state road map, and borehole and well site map. Attribute data used here mainly consists of data on rainfall and existing boreholes and wells abstraction.

TABLE 2: Category of data used

No. Data Year Scale Source
1 SPOT Satellite Image 24 April 2005   MACRES Malaysian Centre for Remote Sensing
2 Rainfall Data 2001-2004   JPS Department of Irrigation and Drainage
3 Borehole and Well Data ( 1987 – 2000 ) 2000   National Water Resources Study ( 2000 – 2050 ) Public Works Department, Department of Irrigation and Drainage, Ministry of Health
4 State Road Map DNMM 9001 2000 1:100,000   JUPEM Department of Surveying and Mapping
5 Topography Map L7030 1996 1:50,000   JUPEM Department of Surveying and Mapping
6 Semenanjung Malaysia -Perlis Map DNMM 9101 2000 1:50,000   JUPEM Department of Surveying and Mapping
7 Geology Map 1985 1:750,000   JMG Department of Mineral and Geoscience

2.2 Darcy’s Law for groundwater flow
To process borehole and well data using GIS, based upon Darcy’s Law for groundwater flow, a 1000m buffer zone around all existing boreholes and wells are considered.
In saturated conditions, one-dimensional flow is governed by Darcy’s Law (Todd,1980) which states that the flow velocity is proportional to the hydraulic gradient

v= k.i
where v = flow velocity
k =coefficient of permeability (measure of soil resistance to flow)
I = hydraulic gradient = H/L

Thus, Whitlow (2001) states that in an aquifer, the quantity of flow through an aquifer is given by

Q= A.v. = A.k.i
where Q =quantity flowing in unit time
A =cross-sectional area through which flow is taking place
I = slope of water table

2.3 Data Processing
Satellite data processing and analysis using ERDAS Imagine software, by method of unsupervised classification of the image, was carried out at Universiti Teknologi MARA (UiTM) Perlis Campus’ Remote Sensing Lab. Next, GIS processing for building the database involved digitizing the topography map, geology map, soil series map and state road map. Using Arc View software, attribute data on volume of daily rainfall was entered to produce a rainfall coverage map. Coordinates (x and y) of existing boreholes and wells were recorded. A landuse map of Perlis state was produced where six classes of landuse were identified as shown in Figure 2; consisting of water bodies (Timah Tasuh dam), human settlement areas, forests, paddy planted areas, sugar cane plantations in Chuping area, and rubber plantations.

FIGURE 2: Thematic Map- Landuse Map of Perlis state

A total of eight thematic layer maps were produced: landuse, geology, rivers and drainage, land administration areas or ‘mukim’, borehole, well, soil types, and rainfall. This follows a similar technique used by previous researchers (Rao Toleti et al,2000; Singh and Prakash,2004). Finally, for analysis using ‘Overlay Method’, a buffer zone of radius 1000m around all existing boreholes and wells were overlaid with each thematic layer map.

FIGURE 3: Thematic Map- Geology Map of Perlis state

FIGURE 4: Thematic Map- Contour Map of Perlis state

3.1 Justification for Choice of Input Data
Satellite data is chosen because previous studies by many other researchers (Khairul Anam et al,2000; Rao Toleti et al,2000; Singh and Prakash,2004). have shown it can provide baseline information on geology/ lithology, structure/ lineaments geomorphology, soil, land cover/ landuse, and hydrological parameters. Other main factors to be considered in implementing groundwater schemes include aquifer characteristics such as Hydraulic Conductivity and Recharge, geological conditions and surrounding land use. Hence the choice of other spatial and attribute data here, consists of topography map, geology map, soil series map, state road map, borehole and well site map, and data on rainfall and boreholes and wells abstraction.

3.2 Result of Analysis using ‘Overlay Method’ on thematic maps
The ‘Overlay Method’ here has been carried out by overlaying the 1000m buffer zone on each individual thematic map. Analysis shows an indication of areas of highest groundwater potential for each thematic layer map. Data integration, which involves combining all the thematic maps have not yet been carried out at this stage. Figure 5 shows the use of the ‘Overlay Method’ where the 1000m buffer zone has been overlaid on the contour map of part of Perlis state.

Analysis on Figure 5 indicates that the majority of existing wells and boreholes are located in the 20 -100 meter contour height, as shown in Table 3 below. Hence an early inference may be made that this region may be of the highest potential.

TABLE 3: Potential groundwater zones in relation to Contour Height

Contour Height (m) Zones
  Numbers %
20-100 49 79.0
100-180 11 17.7
180-260 2 3.3
Total 62 100.00

FIGURE 5: Example of using ‘Overlay Method’ on Contour Map of Perlis state

Further analysis on each of the thematic layers will be done. An analysis on a combination of all layers is expected to provide a more accurate indication on the potential zones of groundwater.

Preliminary results suggests that the spectral and spatial characteristics of SPOT satellite data can serve groundwater monitoring. The results also suggest that a raster-based GIS can facilitate the necessary digital analysis and manipulations, including data integration, geocorrections, and handling classification.

It is foreseen that some of the recommended refinements for future advancement of this study may include addition of other data (e.g. recharge, soil suction, pore pressures, temperature changes) as thematic GIS layers, determining the criterias/ constraint factors, processing of thematic maps using supervised classification, assigning weightage to data layers based on accepted engineering principles to reflect their characteristics and relative importance.

In conclusion, satellite remote sensing and GIS can be used to generate the necessary dynamic information for groundwater monitoring in Perlis state.


  • Aller, L., Bennet, T., Lehr, J.H. and Petty, R.J., 1985. ‘DRASTIC: A standardized system for evaluating groundwater pollution potential using hydrogeologic setting’; EPA/600/2-85/018, R.S.Kerr Environmental Research Laboratory, US EPA, Ada, Oklahoma.
  • Berita Harian, 26 February 2005, ‘Kering’
  • Jones, C.R., 1978 . The geology and mineral resources of Perlis, North Kedah and the Langkawi Island ; Geogical Survey Malaysia District Memoir 17, Kuala Lumpur
  • Kamaruzaman, W.Y., Basharudin, A.H., Che Senu, S., ‘Developing a catchment based resources management system for Perlis towards establishing Perlis GIS: Mapping of landuse cover of Perlis-a preliminary study’, Proceedings ISEG 2005, Malaysia
  • Khairul Anam, Juhari Mat Akhir and Ibrahim Abdullah, 2000. ‘Groundwater prediction potential zone in Langat basin using the integration of RS and GIS’, Proceedings ACRS 2000
  • Rao Toleti, B.V.M. et al, 2000. ‘Integrated groundwater resources mapping in Gurgaon District, Haryana (India) using remote sensing and GIS technique, Proceedings ACRS 2000
  • Singh, A.Kr. and Prakash, S.R., 2004. ‘Integration of thematic maps through GIS for identification of groundwater potential zones’, Proceedings Map India 2004
  • Syed Muhammad, Hooi and Binnie ,2000. ‘Kajian Sumber Air Negara 2000-2050 (Semenanjung Malaysia) State Report (Perlis), Volume 8.
  • The Star, 18 August 2005, ‘Running Dry ‘
  • Tsou, M., Perkins,S.P.,2006. ‘Inverse approaches with lithologic information for a regional groundwater system in southwest Kansas’, Journal of Hydrology, 1 Mar 2006.
  • Todd, D.K., 1980,’Groundwater Hydrology’, 2nd Edition, John Wiley and Sons Inc.
  • Whitlow, R., 2001.’Basic Soil Mechanics’, 4th Edition, Prentice Hall