More reliable and accurate estimates of pollutant loads entering the Great Barrier Reef lagoon
Project Leader: Dr Petra Kuhnert, CSIRO
This MTSRF-funded project has generated a four-step methodology – the Loads Regression Estimator (LRE) - which allows better estimates of load through the calculation of the uncertainty associated with load data. The first step involves ‘regularising’ the flow, a process whereby a less smoother, capturing peak flows, is used to predict flow at regular time intervals (e.g. hourly) and infill any gaps in the flow monitoring data. The second step in the methodology fits a generalised additive model to concentration (on the log-scale) to model the relationship between concentration and a series of flow terms that attempt to mimic key hydrological processes of the system with the aim of reducing knowledge uncertainty. The third step involves the estimation of the load, which is the result of multiplying the regularised flow by the concentration predicted at each regularised flow value, which is then summed over the water year. The fourth step involves the construction of the variance around the estimate of the load. The variance is structured so it not only incorporates errors in the concentration but errors in the flow rates (measurement and spatial). The LRE has been applied to twelve monitored rivers and nine pollutants to estimate annual loads with uncertainties. These estimations were then used to calculate more reliable and accurate long-term loads of pollutants being delivered to the Great Barrier Reef lagoon. The LRE has already been adopted by water quality monitoring programs and is being used to report on the status of, and trends in, pollutant loads in Queensland catchments.
Project 3.7.7 – CSIRO Brown, M. (2007) Final Report, Review of existing approaches used to develop integrated report card frameworks (IRCF) and their relevance to catchments draining to the Great Barrier Reef