Landslide Prediction

Research and development have yielded the know-how and the subsequent software, that can help those countries with tropical rains to predict the likelihood and extent of landslides in areas of highly permeable residual soils.

Often, the countries liable to such heavy rains are those most likely to be involved with large development projects that may be substantially impeded or damaged by resultant landslides.

Such, for example, has been the unfortunate experience with roads important for the economic advancement of Malaysia. There, the cost of countering the collapse of hillside sections has often proven
higher than the original road construction cost.

There is a basic engineering problem in coping with residual soils (those formed in situ by long-term weathering and breakdown of rock) and which are widespread in the Tropics.

In such areas, slope failures are prevalent and are usually set off by marked water precipitation into the highly permeable soils. These may be rapidly infiltrated and are often composed of widely
different component parts. This makes them difficult to sample effectively and to subject them to the necessary shear tests.

During the past 20 years there have been several semi-empirical studies of control of the stability and hydrology of tropical residual soils resulting in a "design by precedent" approach to the
problem. Much of this work was effected in Hong Kong by a study of more than 100 stable and failed slopes when the existing stability-analyses approach seemed ineffective in meeting local conditions.

Subsequent work by Professor Malcolm Anderson and his Department of Geography at the University of Bristol resulted in development of a tensionometer-piezometer designed specifically to cope with the
"worst" soil conditions and to monitor them effectively.

Applying this experience and knowledge garnered during 18 years of earlier work, Professor Anderson and his team developed a program applying mathematical modelling to the flow of water over and
through soils.

Earlier approaches to landslide prediction did not take proper account of how rainfall migrated through soils and affected their dynamic behaviour.

This program, when fed the relevant data resulting from the metering sequences, will calculate changes in the water table of a slope and suggest the angle of slope least likely to result in a
subsequent slide. The program has already been applied beneficially in several countries, including Malaysia, and is contributing markedly to the understanding of what happens in these areas of
terrain instability.

As a result, the researchers developed a coupled-finite difference soil/water model and slope stability analysis for both individual slope and wide area application. This is the software called Chasm
(combined hydrology and stability model) that has been used to redesign the cut of slopes at roadsides in Malaysia.

Chasm has also been used in Hong Kong where much major building development increases the speed and the volume of rainwater run-off and makes the response of the terrain to heavy downfalls complex
and difficult to predict.

Chasm includes in its calculations such multiple influences and their resultant dynamic effects on water percolation and related soil movement. A regional licence for distribution of Chasm has been
negotiated to make it readily available in Malaysia and the Far East.

The fact that slope stability can be increased by trees and shrubs is shown by the fact that following the complete felling of mountain forests there was a higher incidence of landslides, thus
confirming the overall importance of plant-root systems in stabilising hill slopes. This role of vegetation is therefore incorporated within Chasm for Windows.

This "cover" aspect simulates rainfall on a slope with markedly different types of vegetation and so can assess and recommend the best cover to support a particular site.