A comparison of catchment water quality models and direct estimation techniques

Information on water quality, in particular on sediment, phosphorus and nitrogen exports from catchments, is required by catchment managers and other stakeholder groups to determine appropriate management options for water quality issues. However intensive spatial and temporal data on concentrations or loads of these nutrients is relatively scarce. Given this constraint, nutrient export models which can provide useful information on nutrient loads with only sparse data inputs are required. In addition, more complex physics and conceptually based models which are commonly available worldwide are often inappropriate for predicting loads in catchments. Consequently, models of a simpler conceptual and/or empirical type are used where input data are sparse and key processes are catchment dependent. This paper compares four such models (CMSS, AQUALM, Moss model and IHACRESWQ) and several methods based on direct estimation for their ability to predict loads in Australian catchment situations. The implications of the limited available input data sets for load prediction in Australian catchments are discussed. In particular the importance of coordinated data collection over longer time periods and sufficiently fine temporal scales, covering a broad range of event sizes, is emphasised.