Subannual models for catchment management: evaluating model performance on three European catchments.

Models' abilities to predict nutrient losses at subannual timesteps is highly significant for evaluating policy measures, as it enables trends and the frequency of exceedance of water quality thresholds to be predicted. Subannual predictions also permit assessments of seasonality in nutrient concentrations, which are necessary to determine susceptibility to eutrophic conditions and the impact of management practices on water quality. Predictions of subannual concentrations are pertinent to EC Directives, whereas load estimates are relevant to the 50% target reduction in nutrient loading to the maritime area under OSPAR. This article considers the ability of four models (ranging from conceptual to fully mechanistic), to predict river flows, concentrations and loads of nitrogen and phosphorus on a subannual basis in catchments in Norway, England, and Italy. Results demonstrate that model performance deemed satisfactory on an annual basis may conceal considerable divergence in performance when scrutinised on a weekly or monthly basis. In most cases the four models performed satisfactorily, and mismatches between measurements and model predictions were primarily ascribed to the limitations in input data (soils in the Norwegian catchment; weather in the Italian catchment). However, results identified limitations in model conceptualisation associated with the damping and lagging effect of a large lake leading to contrasts in model performance upstream and downstream of this feature in the Norwegian catchment. For SWAT applied to the Norwegian catchment, although flow predictions were reasonable, the large number of parameters requiring identification, and the lack of familiarity with this environment, led to poor predictions of river nutrient concentrations.

Evaluation of the difference of eight model applications to assess diffuse annual nutrient losses from agricultural land.

The capability of eight nutrient models to predict annual nutrient losses (nitrogen and phosphorus) at catchment scale have been studied in the EUROHARP project. The methodologies involved in these models differ profoundly in their complexity, level of process representation and data requirements. This evaluation is focused on model performance in three core catchments: the Vansjø-Hobøl (Norway), the Ouse (Yorkshire, UK) and the Enza (Italy). These three different model applications have been evaluated by comparing calculated annual nutrient loads (total N or nitrate and total P), based on observed flow and total nitrogen or nitrate and total phosphorus concentrations, and the annual nutrient loads that were simulated by the eight nutrient models. Four statistics have been applied for this purpose: the root mean squared error (RMSE), the mean absolute error (MAE), the mean error (ME), and Nash-Sutcliffe's model efficiency (NS). The results show that all model approaches can predict the calculated annual discharges. Depending on the observed statistics (RMSE, MAE, ME and NS) the scores of the model application differed, therefore no overall 'best model' could be identified. Although the water and nutrient loads from (sub)catchments can be predicted, the modelled pathways of nutrients within agricultural land and the nutrient losses to surface waters from agricultural land vary among the catchments and among those model approaches which are able to make this distinction.

Mapping the probability of exceeding critical thresholds for cadmium concentrations in soils in The Netherlands.

The probability of exceeding critical thresholds of Cd concentrations in the soil was mapped at a national scale. The critical thresholds in soil were based on food quality criteria for Cd in crops or in organs of cattle (Bos taurus), and were calculated by inverting a regression model for the Cd concentration in the crop, with the Cd concentration in soil, soil organic matter (SOM) content, clay content, and pH as predictors. The probability of exceeding the critical threshold for Cd in soil per node of a 500- x 500-m grid was approximated by Monte Carlo simulation, using the estimated cumulative distribution functions (cdf) of SOM, clay, pH, and Cd as input. The cdfs were estimated by simple indicator kriging with local prior means. For SOM, clay, and pH, detailed maps of soil type and land use were used to define subregions with assumed constant local means of the indicators (a priori distributions). The cdfs were sampled by Latin hypercube sampling. We accounted for correlation between the actual and critical Cd concentrations in soil by drawing Cd values from cdfs conditional on SOM and clay. The estimated probability for grassland is negligible, even in areas with high Cd concentrations in soil, and for maize (Zea mays L.) land the probability is almost everywhere smaller than 5%. For arable soils, however, these probabilities commonly are larger than 5% when sugar beet (Beta vulgaris L.) or wheat (Triticum aestivum L.) is taken as a reference crop, and locally exceed 50%.