Geoscience and Environment

The Study Area and Model

Penang Island and the Sungai River Basin

Penang Island

The Ara River Basin

The Model

Search Areas

Topography

Topographic Profiles

Go to Sungai Ara Menu

Penang Island

Penang Island, about 300 square kilometers in area, lies in the Indian Ocean about 2.5 km west of the Malaysian mainland. For more than thirty years semiconductor chips have been assembled here, so that almost all electronic devices in the world have components from Penang. The island is a tourist center, famous for its food, the luxury hotels along the north coast, and its sandy beaches. In fact there are only small patches of beach except in the Holocene areas that emerged from the sea about 5,000 years BP. Along much of the coast granite rocks reach to the shore.

Since there are no sedimentary rock strata on the main island, all bedrock is made up of granite that was emplaced in two plutons, divided by a zig-zag line running from Gelugor on a bearing of 290°. Radiometric dating gives and 200±7 MA and 286±10 MA for the northern and southern granites, respectively (Bignell and Snelling, 1977). The highest peaks are just over 825 meters in the north and 450 meters in the south.

Georgetown, the state capital and Malaysia's second largest city, is built on the Quaternary deposits because they are flat and can be excavated for foundations and infrastructure such as water mains, sewers, storm drains, and telephone cables.

The largest area of Holocene and Pleistocene deposits is found in Balik Pulau (literally "back of the island"). Balik Pulau is less urban than the Quaternary areas closer to Georgetown, the bridge, and the ferries to the mainland.

The lower basin of the Sungai Ara was formerly occupied by farmers who cultivated rice in padi fields, but now the basin is partly urbanized.


Go to top of page

Quaternary Geology Penang (GIF 11K)
Scenery (JPG 32K)

The Ara River Basin

The entire lower Ara River basin occupies only about 5 square kilometers. I selected the area for study because it is the right size for a short study, accessible by road, and still has construction activity where sand and gravel are visible at the surface. The view in the photo above is from the northwest corner of the map looking southeastwards.

The map shows that Quaternary deposits, mainly of Holocene age, cover about one-third of the land area (Kamaludin, map, 1992/papers 1989, 1990). The areas designated estuary/lagoon (pale yellow) were probably submerged as sea level rose about 5,000 years ago. The areas marked back mangrove (greenish yellow) and floodplain (yellowish green) were probably always above Holocene sea level, but were either permanently marshy or flooded during the wet season. The topographic map (1977) shows all of these areas as irrigated rice fields.

Today the area is built-up. While the areas designated Pleistocene might have been visually different long ago, today they do not appear any different from the areas marked Holocene. The whole of the Quaternary area is remarkably flat and featureless except for the breaks in slope near the border between the Quaternary and granite.

Quaternary Geology Sungai Ara (GIF 10K)

The granite was probably emplaced at a depth of at least four kilometers (Hutchison, 1977). This can be inferred from the fact that slow cooling at great depth is needed to allow the smaller crystals of microcline to form.

Go to top of page

The Model

I selected oxygen isotope stages 5 and 11 based on the work of the Warm Climates Project of the United States Geological Survey. (Webpage no longer accessible.) The data for the mid-Pliocene is based on (Barron, PRISM), adopting their estimates for sea level high stands. Later I added oxygen isotope stages 7, 9, and 37 from the graph of Nicholas Shackleton, assuming for each that the high stand would be the same as for OIS 5, 5 meters. oisshackelton (GIF 53K)

According to the model, shown in the table below, the amount of uplift during a 100,000-year cycle is estimated as only about 2 meters. This is based an estimate of 76 meters as the present elevation above sea level of the mid-Pliocene high stand (Jones, Walker, and Burton).

This estimate is consistent with the view of Hutchison (1977) that the microcline granites were emplaced at depths of at least four kilmeters, dated radiometrically at 286 MA (Bignell and Snelling, 1977). In Perak the highest peaks rise to only 2,000 meters and generally in Peninsular Malaysia there remain only thin sedimentary strata (Tjia, 1992), thus we could infer total uplift since emplacement of at least 6,000 meters, giving an average rate of uplift of at least 0.02 mm/year, the same order of magnitude as the estimate since the mid-Pliocene. Error would arise from under-estimating the thickness of sedimentary strata over the granite when it was emplaced, under-estimating the depth of emplacement, and under-estimating the rates of erosion and isostatic uplift during the Mesozoic and Tertiary periods, which were warmer and wetter on average than the Pliocene and Pleistocene. However, competent geologists have judged that there has been and continues to be near equilibrium between erosion and isostatic uplift over long periods ((Hall & Nichols, 2002). To establish a working hypothesis I accept this view and the view that the net low rate of uplift results from other causes, such as plate movement. Changes in the geoid could also be a factor (INQUA, N.-A. Mörner).

Model Output

On the basis of the model shown in the following table I will search for evidence of inter-glacial high sea-level stands at 3, 7 to 12, 30, and 75 meters. However, because the lowest map contour is at 15 meters, I will search for high stands at 30 and 75 meters.

Go to top of page

Table (GIF 15K)

Search areas

I adopted a research strategy based on constructing a model to predict the location of sea level high stands at selected time periods. This model suggests elevations for sea level high stands that could possibly marked by terraces.

The map of search areas at the right shows colored bands that define the areas within which the model predicts the location of the sea level high stands for OIS 11 and the mid-Pliocene. The search area map was constructed from a digital elevation model (DEM).

I considered three sources of DEM using:

  • contours extracted from a topographic map; (How to do it.)
  • stereo images, (ASTER)
  • radar data from NASA's Shuttle Radar Topography Mission. ((SRTM).

    Each of these DEMs has advantages and disadvantages. The international SRTM data released is too coarse for this study at 100-meter postings. The ASTER DEM, available with 30-meter postings, is more appropriate. However, the ASTER DEM was acquired in 2002, thus recent earthmoving and building construction masks much of the natural landscape. The DEM constructed from the topographic map surveyed in 1969 contains data that is more likely to conserve terraces formed by earth processes.

    The main handicap in using the DEM generated from a paper topograpic map (DEM-PTM) is the risk that the contour lines themselves will dominate the processing so that they appear as a false signal which might be interpreted as a break in slope indicative of a terrace. The computation process might generate an artifact that appears to be a natural feature.

    • Search Areas (GIF 14K)

    My strategy to deal with this problem follows. I will use the DEM-PTM to guide fieldwork and also to make cross-sectinal profiles. As a check to ensure that not artifacts are being interpreted as natural features I will make a second cross-sectional profile using the ASTER DEM ensuring that the profiles do not cross locations that have been affected by earthmoving or building construction.

    The search areas are located in areas designated as granite, thus there is almost no correspondence between areas shown as search areas and areas marked as Pleistocene on the Quaternary geology map. This is because different approaches have been used to define the areas. The Quaternary geology map shows where deposits of sand, gravel, silt, and organic material have been found, whereas the search areas are hypothetical, based on elevations predicted by the model.

    Go to top of page

    Topography

    The 1969 survey used for the topographic map shows buildings below the 15-meter contour that can be correlated with features shown on the Quaternary map. Topographic map. Above the 15-meter contour there were almost no buildings, thus it is possible to construct profiles of hills to locate terraces. Before constructing profiles I analysed the shape of the basin as a whole by generating a histogram of areas at elevation intervals of one-meter.

    A graph using this data suggests that there are breakpoints in slope near the 45, 60, and 90-meter contour intervals. This is not firm evidence, but it is suggestive of a steeper region between 45 and 60 meters and flatter regions below 45 meters and between 60 and 90 meters. Firmer evidence of terraces must be sought from profiles that cut across the colored bands shown in the search area map.

    Topographic profiles

    Topographic profiles display side views of the landscape much like slices of cake show side views of the cake. The map shows straight lines numbered 1 to 9 that indicate the locations of vertical slices through the landscape. We expect that, if there are terraces, profiles will reveal their shapes.

    The chart showing topographic profiles does suggest that there are terraces at 30 to 35 meters and somewhere between 60 to 90 meters as predicted by the model. I carried out fieldwork to try and resolve the question whether or not there are terraces.

    Go to Fieldwork
    Search profiles (GIF 11K)
    The URL of this site is [www.geoscience2000.info/index.html]