Application of machine-learning models for diagnosing health hazard of nitrate toxicity in shallow aquifers

Agricultural Water Management

Šimůnek

2017

Self Cite

a b s t r a c tNitrate losses are the dominant cause of the non-point source pollution under agricultural fields. In this study, the HYDRUS-2D model was first calibrated and validated using data collected during a two-year field investigation in a drip-irrigated maize field and then applied to evaluate the influence of 176 different N-managed water-saving irrigation scenarios on water and N dynamics and maize grain yield. Various scenarios were defined by combining 11 irrigation levels (IL = 0-100% with a 10% interval), 8 N fertilization rates (NR = 0-400 kg ha −1 with a 50 kg ha −1 interval) and two water-saving irrigation strategies: deficit irrigation (DI) and partial root-zone drying (PRD). Reliable estimates of soil NO 3 − -N concentrations (RMSE = 0.39-10.9 mg l −1 and MBE = −8.9-8.4 mg l −1 ), crop N uptake (RMSE = 3.9-8.9 kg ha −1 and MBE = −5.3-6.25 kg ha −1 ), and soil water contents (RMSE = 2.3-5.11 mm and MBE = 1.63-4.93 mm) were provided by HYDRUS-2D. Based on the simulated results, the fertigation strategy with NR = 200 kg ha −1 is an optimum strategy. For the higher fertigation rates (NR ≥ 250 kg ha −1 ), the NO 3 − -N leaching out of the surface layers (0-20 cm) increased by 0.1-183% while N uptake was enhanced by only 0.3-15%. On the other hand, reducing NR below this level would have resulted in severe economic losses. A 30% reduction in IL at NR = 200 kg ha −1 shows an enormous potential in lowering N leaching below different soil layers (12-99%) while reducing crop N uptake by only 5.4%. In addition, higher crop yield by 0.2-20.2% can be expected under PRD since crop N uptake is enhanced by more water available in the surface layers. While on the one hand, PRD ensures environmentally safer fertilizer applications, on the other hand, the economic objectives are met more easily under PRD than under DI. Additionally, it could be concluded that the HYDRUS-2D model, instead of labor-and time-consuming and expensive field investigations, could be reliably used for determining the optimal scenarios under both the DI and PRD strategies.

Section: Introductionmentioning

confidence: 99%

Two-dimensional modeling of nitrogen and water dynamics for various N-managed water-saving irrigation strategies using HYDRUS

Agricultural Water Management

Šimůnek

2017

Self Cite

“…Efficient models were first determined for each combination of input variables. Then, the various ANFIS models were tested and results obtained were evaluated using different criteria indices (Karandish et al , ).…”

Section: Methodsmentioning

confidence: 99%

Diagnosing Drainage Problems in Coastal Areas Using Machine‐Learning and Geostatistical Models

Darzi‐Naftchali

Irrigation and Drainage

Asgari

2017

Self Cite

This study focuses on diagnosing drainage problems in the coastal areas of Iran by using geostatistical methods, support vector machines (SVMs) and the adaptive neuro‐fuzzy inference system (ANFIS). Groundwater level (WD) and quality were monitored at 37 shallow wells scattered over a 25 000 ha area at different times. Using prepared raster maps of pH, ESP, EC and WD by the best method, drainage problems were categorized into eight classes. Both SVM and ANFIS models significantly improved predicted data for pH, ESP, EC and WD compared with geostatistical models, while SVMs provided slightly better results which were used for further analysis. More than 60% of the area needs drainage to lower the groundwater table in pre‐planting and post‐harvest periods, while during the growing seasons, more than about 72% of the area requires drainage for salinity control. Based on the results, identifying drainage problems at basin scale is possible with cost‐efficient machine learning models with minimum time and data requirements and investment for detailed field surveys. Copyright © 2017 John Wiley & Sons, Ltd.

“…The main reason is that rice grown here requires large amounts of mostly groundwater (cf. Karandish et al, 2017a ). As shown in Table 2 , lower scarcity levels may be achieved when WFs are reduced to the benchmark levels in all climate zones.…”

Section: Groundwater Saving Potentialmentioning

confidence: 99%

“…Agriculture is by far the largest water user: 97% of the total net blue water abstraction in Iran relates to irrigated crop production . The impact of the inefficient agriculture becomes visible through the partial disappearance of lakes, like Urmia Lake ( Ghale et al, 2017 ), the drying up of rivers, like the Zayandehroud River ( Madani, 2014 ), falling groundwater tables ( Rahnema and Mirassi, 2014 ), pollution of water ( Karandish et al, 2017a ), and damage to ecosystems and local livelihoods ( Madani, 2014 ). Moreover, inefficient water use leads to the loss of potential economic benefits that could be derived from increased yields.…”

Section: Case Studymentioning

confidence: 99%

Groundwater saving and quality improvement by reducing water footprints of crops to benchmarks levels

Advances in Water Resources

Hogeboom

2018

Self Cite

a b s t r a c tThe formulation of water footprint (WF) benchmarks in crop production -i.e. identifying reference levels of reasonable amounts of water consumption and pollution per tonne of crop produced -has been suggested as a promising strategy to counter inefficient water use and pollution. The current study is the first to show how setting WF benchmarks may help alleviate groundwater scarcity and pollution, in a case study for Iran. We advance the field of WF assessment by developing WF benchmark levels for crop production, which we successively use to assess potential groundwater saving, quality improvement and economic water productivity gains. First, we calculate climate-specific WF benchmark levels for both total blue water footprints and nitrogen-related grey groundwater footprints for 26 crops, for all years in the period 1980-2010, at 5 × 5 ′ spatial resolution. Second, we estimate the water saving potential for total blue water resources and for groundwater resources specifically, as well as the grey groundwater footprint reduction potential. Finally, we compare mean economic water productivities of crop production in the past with productivities if WFs are reduced to benchmark levels. We find that groundwater comprises up to 83% of total blue water consumption of irrigated crops, with the highest share in arid areas and in cereals. Aquifers are under significant to severe stress, except in the dry sub-humid zone, where irrigation mainly relies on surface water. Reducing WFs of crops to 25th percentile benchmark levels can save 32% of groundwater compared to the reference year 2010, and lower the nitrogen-related grey groundwater footprint by 23%. Moreover, it would increase average economic groundwater productivity in Iran by 20% for cereals, and 59% for nuts. We conclude that reducing WFs to climate-specific benchmark levels in a water-stressed country is a promising way to alleviate overexploitation of aquifers and increase national food security.