Search

High-frequency water quality monitoring is increasingly used in examining the nutrient fluxes within catchments. Despite this, no studies have assessed the impact of monitoring frequency on the uncertainty of nitrate estimates obtained from distributed or semi-distributed catchment models. This study aims to evaluate the impacts of two different frequencies of nitrate sampling on the performance of a catchment hydrology model, including the uncertainty in both predictions and calibrated parameters. The investigation uses the HYPE model to simulate streamflow and nitrate concentrations (2010–2015) in the Selke catchment, a heterogeneous mesoscale catchment in central Germany. The Bayesian inference scheme of the DREAM code was employed for calibration and uncertainty analysis, and to explore differences between fortnightly and daily nitrate sampling strategies. The results indicate that: (a) the posterior uncertainty intervals of nitrogen-export process parameters were narrower when the model was calibrated to daily nitrate measurements, while similar maximum likelihood parameter values were obtained regardless of the sampling frequency; (b) the model calibrated using daily nitrate data better represented both daily and fortnightly nitrate measurements relative to that obtained using fortnightly sampling; (c) the daily nitrate dataset produced significantly smaller parametric prediction uncertainty, but only modest reduction in total prediction uncertainty, relative to the fortnightly nitrate dataset; (d) model structural error and measurement errors are the primary sources of total prediction uncertainty, and these combine to inhibit the benefits of high-frequency monitoring. We conclude that the adequacy of sampling frequency is dependent on model structure and measurement errors, such that higher-frequency nitrate monitoring may not markedly reduce the uncertainty of nutrient predictions, depending on other levels and sources of uncertainty.

Elevated concentration and high variance of dissolved organic carbon (DOC) in surface waters are a challenge for the production of drinking water. Past studies have indicated a dominant role of wetlands in DOC mobilization, but have mainly been focused on boreal and oceanic catchments. Here we analyze the observational DOC time series from 89 temperate humid catchments which drain into German drinking water reservoirs. We characterize the DOC concentration median and variability and utilize partial least squares regression in order to quantify the relation to catchment characteristics such as land use, climate, and topography. We found that the long-term median DOC concentration in the catchment is well predicted by the 90th percentile of the distribution of the topographic wetness index (0.9P TWI) over the entire catchment area. The 0.9P TWI can be directly connected to the abundance of riparian wetlands in the catchments. DOC concentration variability (represented as the ratio of the interquartile range and the median concentrations) was also found to be well predictable. Concentration variability was highest in cold and wet catchments with a high 0.9P TWI. Here we also found stronger correlations between DOC concentrations and discharge, with positive concentration-discharge patterns. Catchments with elevated DOC-concentration variance also exhibited the most severe long-term increases in concentrations. Our results thus indicate that, in temperate climates, riparian wetlands can be the dominant source zones of DOC and control the hydrological mobilization and potentially also the spatial difference in long-term concentration trends observed in surface waters. We conclude that the dominance of topography and climatic conditions in controlling spatio-temporal patterns in DOC concentrations leads to very limited management options.

Summary For capturing spatial variations of runoff and nutrient fluxes attributed to catchment heterogeneity, multi-site hydrological water quality monitoring strategies are increasingly put into practice. This study aimed to investigate the impacts of spatially distributed streamflow and streamwater Inorganic Nitrogen (IN) concentration observations on the identification of a continuous time, spatially semi-distributed and process-based hydrological water quality model HYPE (HYdrological Predictions for the Environment). A Bayesian inference based approach DREAM(ZS) (DiffeRential Evolution Adaptive Metrololis algorithm) was combined with HYPE to implement model optimisation and uncertainty analysis on streamflow and streamwater IN concentration simulations at a nested meso scale catchment in central Germany. To this end, a 10-year period (1994–1999 for calibration and 1999–2004 for validation) was utilised. We compared the parameters’ posterior distributions, modelling performance using the best estimated parameter set and 95% prediction confidence intervals at catchment outlet for the calibration period that were derived from single-site calibration (SSC) and multi-site calibration (MSC) modes. For SSC, streamflow and streamwater IN concentration observations at only the catchment outlet were used. While, for MSC, streamflow and streamwater IN concentration observations from both catchment outlet and two internal sites were considered. Results showed that the uncertainty intervals of hydrological water quality parameters’ posterior distributions estimated from MSC, were narrower than those obtained from SSC. In addition, it was found that the MSC outperformed SSC on streamwater IN concentration simulations at internal sites for both calibration and validation periods, while the influence on streamflow modelling performance was small. This can be explained by the “nested” nature of the catchment and high correlation between discharge observations from different sites. Results revealed, also, that 95% prediction confidence intervals of streamflow and streamwater IN concentration estimated from MSC were more credible compared with those estimated from SSC, which are reflected by narrower confidence intervals and higher percentage of observations bracketed in the estimated unit confidence intervals. The outcomes of this study pointed out the importance of spatially distributed hydrological water quality observations for improving model parameter identification and provided guidelines for choice of adequate calibration strategy and design of hydrological water quality monitoring campaign.