Presentations 2016

Use of early warning monitoring systems for groundwater protection in a policy decision context

Institute for Public Health and the Environment (RIVM)

https://www.youtube.com/watch?v=xJEDDmcZM_I&index=2&list=PLZC-5vj9_JAa9tKW2yjk9Caiz_O8MPKnQ

A recent comparison of monitoring networks showed that sampling methodology strongly affects the measured nitrate concentrations in soil moisture in the vadose zone, affecting the evaluation of Dutch manure policies. The European Nitrates Directive aims at the reduction of nitrate pollution by agriculture. It obliges Member States to abate and prevent groundwater and surface water pollution and to monitor the effectiveness of the actions carried out to comply with the Directive. To ensure that these obligations are met, different early warning monitoring systems have been developed during the last decade. To ensure equal compliance among Dutch soils with large differences in hydro geochemical properties all these systems focus on the assessment of nitrates in soil moisture leaching the rooting zone.The Loess region, in the southern part of the Netherlands, is the smallest of the four soil type regions. Groundwater levels are often deep (>>5 m below surface level) and, therefore, soil moisture is extracted from the soil layer below the root zone to monitor nitrate leaching. Within this region, the national authority (RIVM), the Province of Limburg and the Water Supply Company Limburg (WML) have their own early warning monitoring network. RIVM carries out on-farm monitoring within the entire Loess region as part of the national Minerals Policy Monitoring Programme (LMM). The Province of Limburg carries out a comparable monitoring programme called the ‘Soil Moisture Monitoring’ (BVM). BVM focuses at agricultural fields on the so-called plateaus. These plateaus (higher areas) cover about two-third of the LMM Loess region. The Water Supply Company Limburg is responsible for the project ‘Sustainable Clean Groundwater’ (DSG). DSG aims to ensure the protection of groundwater resources. This project is carried out in co-operation with farmers within the groundwater protection zones.At the beginning of 2014 questions were raised in the agrarian press and the Dutch Parliament about the LMM monitoring protocol for the Loess region. An exploratory research of WML had shown that the WML protocol resulted in 30-40% lower measured nitrate concentrations. As nitrate concentrations in soil moisture are currently not much above the standard of 50 mg/l in the Loess region, this difference between protocols may lead to opposite conclusions. In case of the WML protocol, the conclusion is that almost no additional measures are required, while in case of the LMM protocol, one concludes that still additional measures are needed. WML, the Province of Limburg and RIVM are currently carrying out research – together with other relevant organisations – to determine the extent and cause of this difference. From 2014 onwards, farmers in the Loess region are confronted with additional measures that limit both the use of animal manure and of artificial nitrogen fertiliser to a greater extent than in some other regions. Therefore, the outcome of this research is relevant for farmers in this region. This paper will discuss the scientific, managerial and political dimensions of this research.

Scale dependence of controls on groundwater vulnerability to nonpoint-source nitrate contamination, California Coastal Basin aquifer system

San Francisco State University, Department of Earth & Climate Sciences

https://www.youtube.com/watch?v=uAs8P97oGSc&list=PLZC-5vj9_JAa9tKW2yjk9Caiz_O8MPKnQ&index=1

Understanding the controls on nonpoint-source (NPS) nitrate (NO3–) contamination in groundwater is motivated by the widespread detection of NO3–, implications for human health and aquatic ecosystems, groundwater sustainability, and a growing realization that such understanding across spatial scales promotes management and policy choices that optimize the Water-Energy-Food (WEF) Nexus. In the Coastal California basin aquifer system (CCB), the conflicts in the WEF Nexus are apparent because of intensive agricultural practices that have contributed to chronic NO3– loading to groundwater. Here we evaluate the scale-dependent relations among source, transport, and attenuation (STA) factors that control NPS NO3– contamination in recently recharged CCB groundwater. We present univariate and multivariate logistic regression models for two spatial scales common to management and policy decisions: sub-regional scale (the north, central, and south CCB sub-regions) and regional scale (the entire CCB region). Here we define STA scale dependence if such factors are statistically significant in either sub-regional or regional models and STA scale invariance if such factors are statistically significant in both sub-regional and regional models. We find that dissolved oxygen (DO) (attenuation) in groundwater is often the most significant STA factor in all model domains, indicating that DO is an important, scale-invariant factor controlling NPS NO3– contamination. During the collection of water quality data for this study, we found that DO is not regularly collected during routine groundwater sampling. Considering the importance of DO on denitrification and the relative inexpensive tools used to measure it, DO should be regularly monitored. Farm fertilizer (source) is also a significant scale-invariant factor, while many of the transport factors are scale-dependent factors. We present vulnerability maps that illustrate the spatial patterns of predicted probability of detected elevated NO3–. Findings here improve knowledge about the scale dependence of STA factors, which help decision makers develop best management strategies and policies that advances groundwater sustainability and optimizes the WEF Nexus.

A statistical learning framework for groundwater nitrate models of the Central Valley, California

U.S. Geological Survey

https://www.youtube.com/watch?v=iyKnZL-mAEs&list=PLZC-5vj9_JAa9tKW2yjk9Caiz_O8MPKnQ&index=3

Nitrate is one of the most common anthropogenic contaminants in domestic well water and exceeds the maximum contaminant level of 10 mg/L as N in many wells of the Central Valley, California. Empirical models commonly are used to estimate nitrate potential in groundwater, and can benefit resource managers by identifying the most vulnerable areas. Linear regression and classification methods have been popular choices for estimating nitrate impacts on groundwater. However, such methods can be hampered by hypothesis testing assumptions such as linear and monotonic responses and normally distributed data. Machine learning methods are promising alternatives that dispense with traditional hypothesis testing. For example, tree-based methods do not require data transformation, can fit nonlinear and non-monotonic relations, and automatically incorporate interactions among predictor variables. However, such methods are prone to overfit, which causes the models to not predict well to new data (i.e., samples not used in model training). We used a statistical learning framework to optimize the predictive performance of three machine-learning methods for Central Valley groundwater nitrate data: boosted regression trees (BRT), artificial neural networks (ANN), and Bayesian networks (BN). The statistical learning framework uses cross validation (CV) training and testing data to tune the complexity of the models, and a separate hold-out data set for evaluation of final models. With these data the order of prediction performance based on both CV testing and hold-out R2 values was BRT>BN>ANN. For each method we identified two models based on CV testing results: one with maximum testing R2 and a simpler version with R2 within one standard error of the maximum (the 1SE model). The former yielded training R2 values of 0.94 - 1.0 and the 1SE versions had R2 values of 0.90 - 0.91. Cross-validation testing R2 values indicate predictive performance, and these were 0.22 - 0.39 for the maximum R2 models and 0.19 - 0.36 for the 1SE models. Evaluation with hold-out data suggested that the 1SE BRT and ANN models predicted better to new data (R2 = 0.12 - 0.26) compared with the maximum CV-testing R2 versions. In contrast, a multiple linear regression model explained less than half the variation in the training data and had a hold-out R2 of 0.07. Scatterplots of predicted vs. observed hold-out data obtained for final models helped identify prediction bias, which was greater for ANN and BN than BRT. Spatial patterns of predictions by the final, 1SE BRT model agreed reasonably well with previously observed patterns of nitrate in domestic wells of the Central Valley. According to a map of model predictions, groundwater at domestic well depth in the San Joaquin Valley (south part of the Central Valley) generally is more vulnerable to nitrate than that of the Sacramento Valley to the north, particularly in the eastern and western alluvial fans. Sediments generally are more fine-grained in the Sacramento Valley and have higher Fe and Mn concentrations, conditions that are less conducive to nitrate occurrence in groundwater.

Will our traditions for groundwater sampling in agricultural settings survive the 21st century?

Deltares, Department of Subsurface and Groundwater

https://www.youtube.com/watch?v=ZVuE2ChyCtE&index=4&list=PLZC-5vj9_JAa9tKW2yjk9Caiz_O8MPKnQ

For the evaluation of policy action programs to improve groundwater quality in agricultural catchments, research institutes and governments intensively monitor concentrations of agrochemicals in shallow or near surface groundwater. Conventionally, groundwater is sampled from irrigation or drinking water wells, specially designed groundwater sampling wells or temporary open boreholes. The samples are transported to laboratories and analysed. These conventional procedures are time consuming, expensive, and sometimes unsuitable for the assessment of groundwater quality status and trends. The spatial extent of point sampling from vertical wells and boreholes usually does not align with the scale of our research questions. In addition, the common yearly sampling intervals for upper groundwater do not capture short term weather induced variations in concentrations (Rozemeijer et al, 2009). This causes uncertainty in the groundwater quality status assessment and hampers the detection of trends. Innovative monitoring technologies, like chemical sensors and passive samplers may enhance groundwater quality research. High frequency groundwater quality monitoring using sensors reveals the temporal variability in upper groundwater quality. This reveals the groundwater response to groundwater recharge events as well as diurnal and seasonal cycles. Passive samplers produce time-integrated concentration values at low levels of detection.As long as a detailed assessment of field scale spatial heterogeneity of concentrations in upper groundwater is not the monitoring purpose, a conventional vertical groundwater well may not be to best option. Sampling from a horizontal well, or from a subsurface tile drain that taps the upper groundwater, has proven to yield a spatially field-scale integrated sample of upper groundwater (Rozemeijer et al., 2010). Modern horizontal drilling technology enables the installation of horizontal drains or groundwater monitoring wells at any depth or in any spatial configuration without disturbing the soil structure. Sensors, auto sampers, and passive samplers can be installed at the drain outlet for continuous monitoring and time or flow averaged concentration measurements. Distributed fibre optic sensing has been applied for temperature monitoring at high temporal and spatial resolutions. Current developments in chemical sensing using fibre optic technology may yield detailed distributed groundwater quality data. Innovations in groundwater quality monitoring are retarded due to costs, standardization, and the fear of artificial discontinuities in time series. Groundwater monitoring experimental ‘playgrounds’ and demonstration sites may accelerate the introduction of new technology. After field trials and comparison to conventional monitoring techniques, the application of appropriate innovations can be scaled up. Rozemeijer, J.C., Broers, H.P., Van Geer, F.C. & Bierkens, M.F.P., 2009: Weather-induced temporal variations in nitrateconcentrations in shallow groundwater. – J. Hydrol. 378: 119 –127.Rozemeijer J.C., Van der Velde, Y., Van Geer, F.C., Bierkens, M.F.P. & Broers H.P., 2010: Direct measurements of the tiledrain and groundwater flow route contributions to surface water contamination: From field-scale concentration patternsin groundwater to catchment-scale surface water quality. –Environ. Pollut. 158: 3571– 3579.

Knowledge based protection of groundwater through monitoring and modelling of nitrate in groundwater in rural areas

GEUS

https://www.youtube.com/watch?v=6iw6HLxydNU&index=5&list=PLZC-5vj9_JAa9tKW2yjk9Caiz_O8MPKnQ

Cost efficient sustainable management of groundwater resources in rural areas calls for development of scientifically sound methods usable for mapping and monitoring as wells as dissemination of results to planners and the rural communities.There is a growing awareness in the political and agricultural communities that cost efficient measures to reduce the environmental impact of nitrate and pesticides in the future should be site specific and not only based on overall national regulation. Scientific developments in mapping methods in regard to hydrogeology and geochemistry are a prerequisite before this goal can be realized. Especially a better site specific knowledge of the nitrate reduction in the subsurface and aquifers is required in order to implement site specific measures. The Danish case is a strong example of the potential of the science policy interface to develop better resource efficiency. In Denmark groundwater is synonym with drinking water. To preserve this highly valued resource a national strategy for sustainable abstraction and groundwater protection has developed over the last 50 years. The National Groundwater Mapping Program has detailed mapped the hydro-geo-chemical settings of about 40 % of Denmark followed by delineation of nitrate vulnerable areas in drinking water abstraction areas. Strategic monitoring of the efficiency of implemented measures is part of the cost efficiency. Dating of groundwater has proved indispensable for interpretations. On a national scale regulations implemented by Danish farmers have succeeded in optimizing the nitrogen (N) management at farm level. Accordingly, since the 1980s the overall national upward trend of the nitrate concentrations in oxic groundwater has been reversed (Hansen et al., 2011). Locally, nitrate trend analyses in monitoring wells have shown a more varied pattern with both upward and downward nitrate trends underpinning the need for supplementary site specific groundwater protection initiatives. References:Hansen, B., Thorling, L., Dalgaard, T. & Erlandsen, M., 2011. Trend Reversal of Nitrate in Danish Groundwater – a Reflection of Agricultural Practices and Nitrogen Surpluses since 1950. Environmental Science & Technology, 45, 228-234. Schullehner, J. & Hansen, B., 2014, Nitrate exposure from drinking water in Denmark over the last 35 years. Environmental Research LettersHansen, Birgitte, and Lærke Thorling. 2008. “Use of Geochemistry in Groundwater Vulnerability Mapping in Denmark.” Geological Survey of Denmark and Greenland Bulletin 15: 45–48.Thomsen, R., Søndergaard, V.H. & Sørensen, K.I. 2004: Hydrogeological mapping as a basis for establishing site-specific groundwater protection zones in Denmark. Hydrogeology Journal 12, 550–562.