Sinclair Knight Merz
590 Orrong Road, Armadale
Ph: 03 9248 3303; Fax : 03 9248 3550
An integrated team of leading hydrologists and geospatial professionals at Sinclair Knight Merz has developed GIS techniques to efficiently summarise a suite of catchment parameters. These parameters are numerically analysed and correlated to hydrological data to allow for the assessment and prediction of a variety of catchment conditions. The parameters have been derived for use on a range of projects relating to catchment management issues such as farm dams, stream flow and the impacts of logging throughout New South Wales and Victoria. Prior to the use of digital geospatial information and geographic information systems (GIS), these parameters were derived or estimated manually from hard copy maps where possible. For some parameters, no assessment could be made manually and they could not be taken into consideration.
This paper describes the approach to catchment characterisation and includes an example of the application of catchment characterisation in stream low flow analysis in the Hawkesbury-Nepean Catchment in New South Wales.
There is an increasing wealth of digital geospatial data available for regional use (ie. at scales of 1:25,000 to 1:250,000). This data is being compiled and managed by State and National environmental and land agencies who recognise the importance of geospatial information for environmental management. This data is being valued for its information content as well as for data management and exchange benefits.
The use of Geographic Information Systems (GIS) can open the door for the utilisation of this information in disciplines that are concerned with spatial issues. Without the use of GIS, these disciplines are unable to fully explore spatial relationships in 3D and are restricted to one and two-dimensional analyses. Furthermore, GIS provides a means by which repeatable semi-automated and automated processes can be developed to replace tedious and difficult manual data collection from hard copy maps, diagrams and data bases.
To utilise the current technology and available data, an integrated team of leading hydrologists and geospatial professionals at Sinclair Knight Merz is employing GIS techniques to efficiently summarise a suite of catchment parameters. These parameters are numerically analysed and correlated to hydrological data to allow for the assessment and prediction of a variety of catchment conditions.
The process of catchment characterisation involves obtaining summaries of a number of topographic, climatic and environmental (physiographic) parameters for each catchment in the analysis. The range of parameters to be investigated is limited by the availability of complete and uniform data sets for the region. The uniformity of data sets across all catchments is important to ensure that there are consistent and comparable results.
The first step is to develop catchment or sub-catchment boundary coverages. This can include catchments with water monitoring gauges, nested catchments and ungauged catchments. These catchments may be derived through GIS analysis of digital elevation data (if it is of sufficient quality) or digitised from topographic maps. Each catchment requires a unique identifier.
The available data will vary from state to state and region to region. The core data sets include:
- stream data (generally 1:25,000 available from state land agencies);
- soils (there is a national data set suitable for regional to state wide studies);
- tree cover (generally derived from satellite image analysis);
- DEM (various sources and grid sizes);
- climate data (various sources, including the Bureau of Meteorology); and
- groundwater data sets (various state sources).
The above data sets are available in grid and vector GIS formats and the analyses required to derive the parameters range from vector/grid overlays and intersections to complex stream analysis requiring programs written within the GIS environment. Following the analysis, a number of summarising and reporting functions are required to produce the final catchment characteristics. A sample of typical outputs is shown in Table 1.
Table 1. Sample catchment characterisation data for five parameters
Soil type I/II
Soil type IV
Total stream length (km)
An example of the application of catchment characterisation is the role of GIS in the procedures developed for identifying regions of low flow homogeneity for the specification of environmental flows for the Hawkesbury-Nepean Basin, NSW. A hydrological perspective of the procedures is presented in Nathan et al. (2000) and is summarised below.
Stream flow data is available in the Hawkesbury-Nepean Basin for a number of gauged catchments. However, in order to assess environmental flows for the entire Basin, an extrapolation of this data is required for the “ungauged” catchments. The procedure that was developed by Nathan et al. (1990) involved assigning the ungauged catchments to “indicator” stream flow gauges that can be considered to behave in a similar manner during low flow conditions. The use of GIS analysis, and catchment characterisation, for this purpose allowed a significantly improved methodology to be developed. Firstly by enabling a greater range of physiographic parameters to be taken into account by the analysis, and also by improving the analysis of spatial relationships (topology) between gauging stations, gauged catchments and ungauged catchments.
The GIS tasks for the project are outlined below:
1. The generation of digital boundaries for both the gauged and ungauged catchments, including nested gauged catchments. There were a total of 434 sub-catchments included in the analysis and the boundaries for these were generated by a combined automated and manual analysis of the existing NSW State DEM.
2. The extraction of a large range of catchment characteristics was undertaken for the selected gauged catchments. These were subsequently reduced to six variables found to be the most important in explaining the variation in recorded stream flow.
3. Catchment characteristics for the six variables were then obtained for all catchments in the Hawkesbury-Nepean Basin. These results were analysed using cluster analysis to identify homogenous groupings of catchments. There were four distinct classes derived reflecting hydrological regions of low flow response. The final assignment of ungauged catchments to gauges followed a weighted function based on hydrologic similarity and geographic proximity.
GIS is an extremely useful tool for environmental management. The current digital geospatial and image data available in Australia enables spatial analysis to be undertaken on a range of regional and catchment wide projects. Integrated teams of environmental scientists, engineers and geospatial professionals are able to fully realise the value of these existing data sets and to effectively use GIS at this scale.
Catchment characterisation is a good example of the use of GIS as it has enabled the:
- access to and utilisation of a broad range of available digital environmental data,
- development of new methodologies to better resolve fundamental hydrological issues, and
- results are required at a scale that is compatible with the current suite of available data.
Nathan, R.J, Rahman, A., Bagg, S. and Green, J.H. (2000). An objective procedure for identifying regions of low flow homogeneity for the specification of environmental flows. Paper submitted to 3rd International Hydrology and Water Resources Symposium, November 2000.