Influence of different nitrate-N monitoring strategies on load estimation as a base for model calibration and evaluation

Effects of sampling strategy in rivers on load estimation for Nitrate-Nitrogen and total Phosphorus in a lowland agricultural area

The transport of nutrients into water bodies is one of the main causes of water eutrophication. It is therefore important to estimate the loads of nutrients. Discharge and nutrient concentrations are the fundamental elements to estimate the loads of nutrients, the latter can be affected by sampling strategies. As conducting sampling campaign and laboratory analysis are both expensive, it is necessary to find the best effective sampling strategy. The aim of this paper is to show how autocorrelation and standard statistical methods can be used to test the effects of different sampling strategies on the nutrient load estimation and to find the optimal sampling strategy. The data set in this study is from the 50 km² Kielstau catchment, a UNESCO demo site for ecohydrology in Northern Germany and consists of 14 years daily values of climate, hydrology, and water quality from 2006 to 2019. We calculated the autocorrelation (AC) of discharge (Q), precipitation, Nitrate-Nitrogen (NO₃-N) and total Phosphorus (Ptot). Then we tested the effects of sampling intervals from 7 to 56 days (1-8 weeks) on the nutrient loads. Our results showed a high AC of Q and NO₃-N for a long period, but the AC of Ptot and precipitation decreased very fast. An increase of the sampling interval (less frequent) increased the error of estimating the concentrations and loads. Consequently, we recommend that (1) the optimal sampling strategy for nutrient load estimation in an agriculture-dominant catchment should be continuously monitoring discharge combined with periodic grabbed samples; (2) the sampling frequency for NO₃-N is suggested to be monthly (every 28 days) and for Ptot weekly (every 7 days). The information will help those tasked with catchment monitoring to design appropriate sampling strategy to ensure adequate data for nutrients load estimation in lowland rivers.

Reservoir water quality deterioration due to deforestation emphasizes the indirect effects of global change

Deforestation is currently a widespread phenomenon and a growing environmental concern in the era of rapid climate change. In temperate regions, it is challenging to quantify the impacts of deforestation on the catchment dynamics and downstream aquatic ecosystems such as reservoirs and disentangle these from direct climate change impacts, let alone project future changes to inform management. Here, we tackled this issue by investigating a unique catchment-reservoir system with two reservoirs in distinct trophic states (meso‑ and eutrophic), both of which drain into the largest drinking water reservoir in Germany. Due to the prolonged droughts in 2015-2018, the catchment of the mesotrophic reservoir lost an unprecedented area of forest (exponential increase since 2015 and ca. 17.1% loss in 2020 alone). We coupled catchment nutrient exports (HYPE) and reservoir ecosystem dynamics (GOTM-WET) models using a process-based modeling approach. The coupled model was validated with datasets spanning periods of rapid deforestation, which makes our future projections highly robust. Results show that in a short-term time scale (by 2035), increasing nutrient flux from the catchment due to vast deforestation (80% loss) can turn the mesotrophic reservoir into a eutrophic state as its counterpart. Our results emphasize the more prominent impacts of deforestation than the direct impact of climate warming in impairment of water quality and ecological services to downstream aquatic ecosystems. Therefore, we propose to evaluate the impact of climate change on temperate reservoirs by incorporating a time scale-dependent context, highlighting the indirect impact of deforestation in the short-term scale. In the long-term scale (e.g. to 2100), a guiding hypothesis for future research may be that indirect effects (e.g., as mediated by catchment dynamics) are as important as the direct effects of climate warming on aquatic ecosystems.

Comparison of Long Short-Term Memory and Weighted Regressions on Time, Discharge, and Season Models for Nitrate-N Load Estimation

The long short-term memory (LSTM) model has been widely used for a broad range of applications entailing the estimation of variables in different fields to improve water quality management in rivers. The main objectives of this study are (1) to develop a novel LSTM-based model for the estimation of nitrate-N loads, which adversely affect water resources, and (2) to evaluate the performance of the model by comparing it with that of Monte Carlo sub-sampling and the weighted regressions on time discharge and season (WRTDS) model. We evaluated the model performance using various numbers of hidden layers, ranging from one to four, in the LSTM model to determine the appropriate number of hidden layers; furthermore, we applied the sampling frequencies of 6, 12, and 24 to assess their impact. Seven polluted river basins in the United States were used for analysis, and the relative root mean squared error (rRMSE) and the mean percentage error (MPE) metrics were applied for the validation of the model estimates. The proposed model achieved accurate nitrate-N load estimates using three to four hidden layers, and improved model performance was observed when the sampling frequency was increased. The differences among the results obtained using the LSTM model were examined based on a binning technique via a log-log plot of nitrate-N concentration against discharge. The binning analysis showed that the slope obtained from the average rates of discharge and low discharge values apparently influenced the estimates. Furthermore, box plot analyses of the statistical indices such as rRMSE and MPE demonstrate that the LSTM model seems to exhibit better performance than the WRTDS model. The results of the examination demonstrate that the LSTM model may be a good alternative with regard to estimating nitrate-N loads for the control of water quality constituents.

Nutrient pollutant loading and source apportionment along a Mediterranean river

Rivers are increasingly being subjected to increased anthropogenic pollution stresses that undermine their designated uses and negatively affect sensitive coastal regions. The degradation of river water quality is attributed to both point and nonpoint sources of pollution. In this study, we determine the relative contribution of point and nonpoint pollutant loads in the Beirut River basin, a poorly monitored seasonal Mediterranean river. Water quality samples were collected on a weekly basis over 2 consecutive years (2016 and 2017) from four sampling sites that represent a gradient of increasing urbanization. Flow-concentration models were first developed to estimate total phosphorus (TP), total nitrogen (TN), and total suspended solids (TSS) loads reaching the different sub-basins. The performance of the regression models varied by location and by pollutant, with improved performance in the downstream sections (adjusted R² 66% for TP and 59% for TN). Loads were also determined using the Beale's ratio method, which generally underestimated the loads as compared with the regression-based models. The relative contribution of the nonpoint source loads were then quantified using the Open Nonpoint Source Pollution and Erosion Comparison Tool (OpenNSPECT). The results showed that point sources were the main cause of water quality impairment across the entire basin, with load contributions varying between 75% in the headwaters and 98% in the urbanized downstream sections. The adopted modeling approach in this study provides an opportunity to better understand pollutant load dynamics in poorly monitored basins and a mechanism to apportion pollution loads between point and nonpoint sources.

Influence of sampling frequency and load calculation methods on quantification of annual river nutrient and suspended solids loads

Better management of water quality in streams, rivers and lakes requires precise and accurate estimates of different contaminant loads. We assessed four sampling frequencies (2 days, weekly, fortnightly and monthly) and five load calculation methods (global mean (GM), rating curve (RC), ratio estimator (RE), flow-stratified (FS) and flow-weighted (FW)) to quantify loads of nitrate-nitrogen (NO₃^--N), soluble inorganic nitrogen (SIN), total nitrogen (TN), dissolved reactive phosphorus (DRP), total phosphorus (TP) and total suspended solids (TSS), in the Manawatu River, New Zealand. The estimated annual river loads were compared to the reference 'true' loads, calculated using daily measurements of flow and water quality from May 2010 to April 2011, to quantify bias (i.e. accuracy) and root mean square error 'RMSE' (i.e. accuracy and precision). The GM method resulted into relatively higher RMSE values and a consistent negative bias (i.e. underestimation) in estimates of annual river loads across all sampling frequencies. The RC method resulted in the lowest RMSE for TN, TP and TSS at monthly sampling frequency. Yet, RC highly overestimated the loads for parameters that showed dilution effect such as NO₃^--N and SIN. The FW and RE methods gave similar results, and there was no essential improvement in using RE over FW. In general, FW and RE performed better than FS in terms of bias, but FS performed slightly better than FW and RE in terms of RMSE for most of the water quality parameters (DRP, TP, TN and TSS) using a monthly sampling frequency. We found no significant decrease in RMSE values for estimates of NO₃^-N, SIN, TN and DRP loads when the sampling frequency was increased from monthly to fortnightly. The bias and RMSE values in estimates of TP and TSS loads (estimated by FW, RE and FS), however, showed a significant decrease in the case of weekly or 2-day sampling. This suggests potential for a higher sampling frequency during flow peaks for more precise and accurate estimates of annual river loads for TP and TSS, in the study river and other similar conditions.

Seasonal variation, flux estimation, and source analysis of dissolved emerging organic contaminants in the Yangtze Estuary, China

The occurrence and seasonal variation of 24 dissolved emerging organic contaminants in the Yangtze Estuary were studied, including 12 non-antibiotic pharmaceuticals, seven sulfonamides, two macrolides and three chloramphenicols. Sulfadiazine, erythromycin, thiamphenicol and paracetamol were the primary contaminants in sulfonamides, macrolides, chloramphenicols and non-antibiotic pharmaceutical groups, respectively. Compared to the concentrations at Datong, chloramphenicols at Xuliujing were significantly higher in autumn and winter, while macrolides were lower in spring. Based on the flux estimation, approximately 37.1 tons of sulfonamides, 17.4 tons of macrolides, 79.2 tons of chloramphenicols and 14.1 tons of non-antibiotic pharmaceuticals were discharged into the Yangtze Estuary from June 2013 to May 2014. However, the total flux from the Huangpu River only represented 5% of the total. The pharmaceutical sources were speculated on by analyzing the seasonal variations in pharmaceutical concentrations and fluxes at various sites. Both environmental and social factors might affect the fluxes.

Modeling Multi-Event Non-Point Source Pollution in a Data-Scarce Catchment Using ANN and Entropy Analysis

Event-based runoff–pollutant relationships have been the key for water quality management, but the scarcity of measured data results in poor model performance, especially for multiple rainfall events. In this study, a new framework was proposed for event-based non-point source (NPS) prediction and evaluation. The artificial neural network (ANN) was used to extend the runoff–pollutant relationship from complete data events to other data-scarce events. The interpolation method was then used to solve the problem of tail deviation in the simulated pollutographs. In addition, the entropy method was utilized to train the ANN for comprehensive evaluations. A case study was performed in the Three Gorges Reservoir Region, China. Results showed that the ANN performed well in the NPS simulation, especially for light rainfall events, and the phosphorus predictions were always more accurate than the nitrogen predictions under scarce data conditions. In addition, peak pollutant data scarcity had a significant impact on the model performance. Furthermore, these traditional indicators would lead to certain information loss during the model evaluation, but the entropy weighting method could provide a more accurate model evaluation. These results would be valuable for monitoring schemes and the quantitation of event-based NPS pollution, especially in data-poor catchments.

Quantification of Phosphorus Exports from a Small Forested Headwater-Catchment in the Eastern Ore Mountains, Germany

Phosphorus (P) export from forest soils is mainly driven by storm events, which induce rapid flow processes by preferential flow bypassing large parts of the soil matrix. However, little is known about the dynamics, magnitude, and driving processes of P exports into surface waters. In this paper, we present the results of a monitoring study in a small forested catchment (21 ha) situated in the low mountain ranges of Saxony, Germany. During the fixed schedule-sampling (weekly to bi-weekly sampling frequency for a three-year period), a mean total-P concentration of 8 μg·L−1 was measured. However, concentrations increased up to 203 μg·L−1 during individual storm flow events. Based on the analyzed concentrations and continuously measured discharge we calculated mean annual export rates of 19 to 44 g·ha−1·a−1 for the weekly sampling frequency with different load calculation methods. If events are included into the annual load calculation, the mean annual export fluxes can be up to 83 g·ha−1·a−1 based on the different load calculation methods. Predictions of total-P export rates based on a sampling strategy which does not consider short-term changes due to factors such as storms will substantially underestimate P exports.

Impacts of Watershed Characteristics and Crop Rotations on Winter Cover Crop Nitrate-Nitrogen Uptake Capacity within Agricultural Watersheds in the Chesapeake Bay Region

The adoption rate of winter cover crops (WCCs) as an effective conservation management practice to help reduce agricultural nutrient loads in the Chesapeake Bay (CB) is increasing. However, the WCC potential for water quality improvement has not been fully realized at the watershed scale. This study was conducted to evaluate the long-term impact of WCCs on hydrology and NO3-N loads in two adjacent watersheds and to identify key management factors that affect the effectiveness of WCCs using the Soil and Water Assessment Tool (SWAT) and statistical methods. Simulation results indicated that WCCs are effective for reducing NO3-N loads and their performance varied based on planting date, species, soil characteristics, and crop rotations. Early-planted WCCs outperformed late-planted WCCs on the reduction of NO3-N loads and early-planted rye (RE) reduced NO3-N loads by ~49.3% compared to the baseline (no WCC). The WCCs were more effective in a watershed dominated by well-drained soils with increased reductions in NO3-N fluxes of ~2.5 kg N·ha-1 delivered to streams and ~10.1 kg N·ha-1 leached into groundwater compared to poorly-drained soils. Well-drained agricultural lands had higher transport of NO3-N in the soil profile and groundwater due to increased N leaching. Poorly-drained agricultural lands had lower NO3-N due to extensive drainage ditches and anaerobic soil conditions promoting denitrification. The performance of WCCs varied by crop rotations (i.e., continuous corn and corn-soybean), with increased N uptake following soybean crops due to the increased soil mineral N availability by mineralization of soybean residue compared to corn residue. The WCCs can reduce N leaching where baseline NO3-N loads are high in well-drained soils and/or when residual and mineralized N availability is high due to the cropping practices. The findings suggested that WCC implementation plans should be established in watersheds according to local edaphic and agronomic characteristics for reducing N leaching.

Assessment of the Spatial and Temporal Variations of Water Quality for Agricultural Lands with Crop Rotation in China by Using a HYPE Model

Many water quality models have been successfully used worldwide to predict nutrient losses from anthropogenically impacted catchments, but hydrological and nutrient simulations with limited data are difficult considering the transfer of model parameters and complication of model calibration and validation. This study aims: (i) to assess the performance capabilities of a new and relatively more advantageous model, namely, Hydrological Predictions for the Environment (HYPE), that simulates stream flow and nutrient load in agricultural areas by using a multi-site and multi-objective parameter calibration method and (ii) to investigate the temporal and spatial variations of total nitrogen (TN) and total phosphorous (TP) concentrations and loads with crop rotation by using the model for the first time. A parameter estimation tool (PEST) was used to calibrate parameters. Results show that the parameters related to the effective soil porosity were highly sensitive to hydrological modeling. N balance was largely controlled by soil denitrification processes. P balance was influenced by the sedimentation rate and production/decay of P in rivers and lakes. The model reproduced the temporal and spatial variations of discharge and TN/TP relatively well in both calibration (2006–2008) and validation (2009–2010) periods. Among the obtained data, the lowest Nash-Suttclife efficiency of discharge, daily TN load, and daily TP load were 0.74, 0.51, and 0.54, respectively. The seasonal variations of daily TN concentrations in the entire simulation period were insufficient, indicated that crop rotation changed the timing and amount of N output. Monthly TN and TP simulation yields revealed that nutrient outputs were abundant in summer in terms of the corresponding discharge. The area-weighted TN and TP load annual yields in five years showed that nutrient loads were extremely high along Hong and Ru rivers, especially in agricultural lands.

A modeling approach for agricultural water management in citrus orchards: cost-effective irrigation scheduling and agrochemical transport simulation

The water flow and the mass transport of agrochemicals in the unsaturated and saturated zone were simulated in the extended alluvial basin of Keritis river in Crete, Greece (a predominantly flat and most productive citrus growing area) using the hydrological model MIKE SHE. This model was set up based on information on land use, geology, soil structure, meteorological data, as well as groundwater level data from pumping wells. Additionally, field measurements of the soil moisture at six different locations from three soil depths (0.1, 0.2, and 0.3 m) were used as targets to calibrate and validate the unsaturated flow model while for saturated condition, groundwater level data from three well locations were used. Following the modeling approach, the agrochemical mass transport simulation was performed as well, based on different application doses. After the successful calibration processes, the obtained 1D modeling results of soil moisture-pressure related to soil depth at different locations were used to design a proper and cost-effective irrigation programme (irrigation timing, frequency, application rates, etc.) for citrus orchards. The results of the present simulation showed a very good correlation with the field measurements. Based on these results, a proper irrigation plan can be designed at every site of the model domain reducing the water consumption up to 38% with respect to the common irrigation practices and ensuring the citrus water productivity. In addition, the effect of the proposed irrigation scheduling on citrus yield was investigated. Regarding the agrochemical concentration in the groundwater for all dose cases was below the maximum permissible limit. The only exception was for the highest dose in areas where the water table is high. Thus, this modeling approach could be used as a tool for appropriate water management in an agricultural area estimating at each time and location the availability of soil water, contributing to a cost-effective irrigation plan.

SWAT meta-modeling as support of the management scenario analysis in large watersheds

In the last two decades, numerous models and modeling techniques have been developed to simulate nonpoint source pollution effects. Most models simulate the hydrological, chemical, and physical processes involved in the entrainment and transport of sediment, nutrients, and pesticides. Very often these models require a distributed modeling approach and are limited in scope by the requirement of homogeneity and by the need to manipulate extensive data sets. Physically based models are extensively used in this field as a decision support for managing the nonpoint source emissions. A common characteristic of this type of model is a demanding input of several state variables that makes the calibration and effort-costing in implementing any simulation scenario more difficult. In this study the USDA Soil and Water Assessment Tool (SWAT) was used to model the Venice Lagoon Watershed (VLW), Northern Italy. A Multi-Layer Perceptron (MLP) network was trained on SWAT simulations and used as a meta-model for scenario analysis. The MLP meta-model was successfully trained and showed an overall accuracy higher than 70% both on the training and on the evaluation set, allowing a significant simplification in conducting scenario analysis.

Saving the Baltic Sea, the inland waters of its drainage basin, or both? spatial perspectives on reducing P-loads in eastern Sweden

Nutrient loads from inland sources to the Baltic Sea and adjacent inland waters need to be reduced in order to prevent eutrophication and meet requirements of the European Water Framework Directive (WFD) and the Baltic Sea Action Plan (BSAP). We here investigate the spatial implications of using different possible criteria for reducing water-borne phosphorous (P) loads in the Northern Baltic Sea River Basin District (NBS-RBD) in Sweden. Results show that most catchments that have a high degree of internal eutrophication do not express high export of P from their outlets. Furthermore, due to lake retention, lake catchments with high P-loads per agricultural area (which is potentially of concern for the WFD) did not considerably contribute to the P-loading of the Baltic Sea. Spatially uniform water quality goals may, therefore, not be effective in NBS-RBD, emphasizing more generally the need for regional adaptation of WFD and BSAP-related goals.

Land use change impacts on water quality in three lake winnipeg watersheds

Lake Winnipeg eutrophication results from excess nutrient loading due to agricultural activities across the watershed. Estimating nonpoint-source pollution and the mitigation effects of beneficial management practices (BMPs) is an important step in protecting the water quality of streams and receiving waters. The use of computer models to systematically compare different landscapes and agricultural systems across the Red-Assiniboine basin has not been attempted at watersheds of this size in Manitoba. In this study, the Soil and Water Assessment Tool was applied and calibrated for three pilot watersheds of the Lake Winnipeg basin. Monthly flow calibration yielded overall satisfactory Nash-Sutcliffe efficiency (NSE), with values above 0.7 for all simulations. Total phosphorus (TP) calibration NSE ranged from 0.64 to 0.76, total N (TN) ranged from 0.22 to 0.75, and total suspended solids (TSS) ranged from 0.29 to 0.68. Based on the assessment of the TP exceedance levels from 1993 to 2007, annual loads were above proposed objectives for the three watersheds more than half of the time. Four BMP scenarios based on land use changes were studied in the watersheds: annual cropland to hay land (ACHL), wetland restoration (WR), marginal annual cropland conversion to hay land (MACHL), and wetland restoration on marginal cropland (WRMAC). Of these land use change scenarios, ACHL had the greatest impact: TSS loads were reduced by 33 to 65%, TN by 58 to 82%, and TP by 38 to 72% over the simulation period. By analyzing unit area and percentage of load reduction, the results indicate that the WR and WRMAC scenarios had a significant impact on water quality in high loading zones in the three watersheds. Such reductions of sediment, N, and P are possible through land use change scenarios, suggesting that land conservation should be a key component of any Lake Winnipeg restoration strategy.

The impact of Best Management Practices on simulated streamflow and sediment load in a Central Brazilian catchment

The intense use of water for both public supply and agricultural production causes societal conflicts and environmental problems in the Brazilian Federal District. A serious consequence of this is nonpoint source pollution which leads to increasing water treatment costs. Hence, this study investigates in how far agricultural Best Management Practices (BMPs) might contribute to sustainable water resources management and soil protection in the region. The Soil and Water Assessment Tool (SWAT) was used to study the impact of those practices on streamflow and sediment load in the intensively cropped catchment of the Pipiripau River. The model was calibrated and validated against measured streamflow and turbidity-derived sediment loads. By means of scenario simulations, it was found that structural BMPs such as parallel terraces and small sediment basins ('Barraginhas') can lead to sediment load reductions of up to 40%. The implementation of these measures did not adversely affect the water yield. In contrast, multi-diverse crop rotations including irrigated dry season crops were found to be disadvantageous in terms of water availability by significantly reducing streamflow during low flow periods. The study considers rainfall uncertainty by using a precipitation data ensemble, but nevertheless highlights the importance of well established monitoring systems due to related shortcomings in model calibration. Despite the existing uncertainties, the model results are useful for water resource managers to develop water and soil protection strategies for the Pipiripau River Basin and for watersheds with similar characteristics.