Modelling aids Malaysian flood control
River system modelling and flood forecasting are growing in importance in Asian water management. Abd Jalil Hassan and Md Nassir Md Noh of the Department of Irrigation and Drainage at the National Hydraulic Research Institute in Malaysia describe a recent project in the Selangor basin.
Damage caused by flooding is estimated at around US$264 million in tangible costs and a further US$927 million in intangible costs. For this reason, the Malaysian government has decided on a number of measures:
- The development of a comprehensive river model consisting of hydrological, hydraulic and ground elements, covering the country's main river and flood plain.
- Production of a flood risk map based on various return periods, and to produce quick results regarding flood impact due to likely human activities in a catchment. The aim was to model the whole river system as one unit, and to explore and explain the effects of varying runoff rates from the sub-catchments, as well as eliminating the existing dependence on rating curves (a curve predicting water height from the volume of flow in a river).
- To model the impact of flooding caused by changes in catchment characteristics.
The study area is the basin of the Selangor river, a 1960km2 catchment area whose other contributory rivers include the Sembah, Kerling, Batang Kali, Rening and Buloh. The Selangor itself is 110km long.
Critical hydrological data acquired for the modelling exercise included rainfall, water level and flow data from the Department of Irrigation and Drainage (DID) and land use and soil type data at 1:50,000 scale from the Department of Agriculture.
Hydraulic data included 1:2000 river cross-sections and structure information from DID, as well as tidal data from Malaysia's survey and mapping organisation - JUPEM - and the navy. Ground data came from 1:5000 3D flood-plain terrain data, again provided by DID.
The Malaysian government chose to set up the model using UK technology company Wallingford Software's InfoWorks RS. The Selangor river was modelled using cross-sections nearly 1km apart, which were extended to the flood plain at between 0.5km and 1.1km apart.
Subcatchments were represented as hydrological units at their respective locations using the US standard method, USSCS. Flood plains were represented as extended sections and reservoir units, connected by spill units.
The model was calibrated for two major flooding events that took place in 1971 and 1988. The Manning roughness coefficient used in the model was 0.04 for the main channel and 0.1 to 0.2 for the flood plain. The spill unit coefficient used was 1.0 for river overflow and 0.1 to 0.5 for the flood plain.
To model the impact of development, some assumptions were made: that the river Sembah sub-catchment land-use was fully developed, the SCS Curve Number was from 65 to 80, TC (time of concentration) was from 31.4 hours to 20.6 hours, and that the comparator would be a 100-year return period.
Wallingford Software's products include data management and network modelling software to support planning and operations in water distribution, sewerage, river management and coastal engineering. Customers include water authorities, consulting engineers and academic institutions throughout the world.
Wallingford's core solutions comprise:
- InfoWorks: a suite of three integrated yet separate hydraulic modelling software solutions that embrace the whole water cycle from supply and distribution (InfoWorks WS), urban drainage and wastewater management (InfoWorks CS) through to river and coastal management (InfoWorks RS)
- FloodWorks: offering real time flood forecasting, monitoring, event notification, and decision support for operational control rooms.
- InfoNet: an asset data management system for water and wastewater networks that enables users to integrate, validate, clean, control, store and report upon the wealth of diverse data that exists within water organisations from GIS data to CCTV images.