Decision Support Tools for Managing Water Quality
Harvey JR Rodda, NIWA, Hamilton
The term Decision Support System has become popular in recent years in the fields of hydrology and environmental management to describe user friend]y computer packages which are aimed at providing understandable advice for the end user. Decision Support Systems are the progressive development of computer models, which were formerly used solely for research purposes, into management tools which can be used without requiring a detailed understanding of the subject in question or computing Models could be any array of algorithms which were capable of predicting a specific process, whereas a Decision Support System combines a model with a user interface so that the required data can be readily input, processed, and then output in a form which is easy to understand. The output in particular is designed to assist the user with management options rather than give a definite answer, hence the term 'decision support'.
The work being undertaken by NIWA and AgResearch will combine a model for predicting water quality (BNZ), and an agricultural production model (OUTLOOK) with an existing user interface generated by the ARC/INFO GIS.
The GIS interface will allow the access to three countrywide databases describing land use (Land Resources
Inventory), hydrology and water resources (Water Resources Archive) and climate (National Climate Database).
These databases have recently been generated in ARC/INFO by NIWA in Christchurch. Information is in the form of maps using 1km by lkm pixels, l00m contour lines, catchment boundaries, river channel networks and climate stations (Shankar, 1995) The databases can be used to generate hydrographs, and a model is currently in use to predict diurnal temperature for a given location on the map.
The modelling basis for the Decision Support System will be provided by the BNZ (Basin New Zealand) and OUTLOOK models. BNZ predicts the runoff, sediment and nutrient losses on a catchment scale (Cooper et al, 1992), and OUTLOOK compares and evaluates the economics of different fertiliser strategies (Metherell et al., 1995). The BNZ model requires input data concerning land use, soil type, topography, fertiliser applications, rainfall, temperature and solar radiation. The amount and diversity of information held within the above databases could easily be used to generate this model input data Also the use of the ARC/INFO GIS will greatly simplify a number of the operations required for running BNZ. For example, BNZ requires a network of grid cells to be created for the catchment, with input data taken from each cell. By using ARC/INFO, such operations can be completed in a matter of seconds. The output from BNZ could be spatially represented by ARC/INFO in a similar way to the information held within the databases (i.e., using maps and graphs).
The combination of BNZ and these databases within an ARC/INFO interface would provide a decision support tool for predicting the spatial and temporal variations of runoff, sediment and nutrient losses for any agricultural catchment listed on the database. The effects of land use change and different fertiliser application rates on water quality could also be examined. To address the environmental effects of different farm strategies, the economics oriented OUTLOOK model needs to be included into the system. Different fertiliser and stocking policies can be examined using OUTLOOK, on the basis of the returns predicted. New values for the nutrient status of the land in question will be calculated which in turn will be fed back into the BNZ model generating a new output in terms of water quality. OUTLOOK will also be modified to produce more general outputs of farm production (not just grazed pastures).
These are currently the Toenepi and Scotsmans Valley catchments near Hamilton and the Ngongotaha at Rotorua. Further catchments may be available for testing in the future. BNZ has already been tested successfully for the Scotsmans Valley catchment.
The Decision Support System will also have the facility to incorporate other environmental models. For example, models predicting the effects of forestry on water quality and soil erosion could be included, New modelling approaches such as the recently developed MMS Modular Modelling System) from the USA (Flugel, 1995) could be interchanged with the existing models, when they become available. In this way, the Decision Support System will be able to keep pace with new technology.
References
Cooper, A B, Smith, CM, and Bottcher, A B, (1992). Predicting runoff of water, sediment, and nutrients from a New Zealand grazed pasture using CREAMS. Trans. Am. Soc Ag. Engrs. Vol 35, No.1,105-112.
Flugel, W A, (1995). Hydrological response units (HRUs) to preserve basis heterogeneity in hydrological modelling using PRMS/MMS - case study in the Brol basin, Germany. In Simonovic, S P, Kundzewicz, Z, Rosbjerg, D, and Takeuchi, K (Eds) Modelling and Management of Sustainable Basin Scale Water Resource Systems. IAHS Publication No.231, Wallingford, UK, 79-87.
Metherell, A K, McCall, D O, and Woodward, S R 1, (1995). OUTLOOK™: a phosphorus fertiliser decision support model for grazed pastures. Fertiliser Requirements of Grazed Pasture and Field Crops. Fertiliser and Lime Research Centre Workshop Proceedings, Massey University, 24-39.
Shankar, U, (1995). GIS: an aid to environmental modelling. Water and Atmosphere, Vol.3, No.2, NIWA, Christchurch, 24-25.
Contact for Enquiries
MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND
Fax: +64 4 894 0721
Contact this person
