Stochastic nitrate simulation under hydraulic conductivity uncertainty of an agricultural basin aquifer at Eastern Thessaly, Greece

Environ Sci Pollut Res Int. 2021 Dec;28(46):65700-65715. doi: 10.1007/s11356-021-15555-1. Epub 2021 Jul 28.

Abstract

Lake Karla Watershed is an agricultural basin characterized by intense agricultural activities, which lead to quantitative and qualitative aquifer degradation. The exploitation of non-renewable groundwater resources and nitrate contamination are the major threats to aquifer sustainability. Groundwater resources cover mainly irrigation and domestic water needs. Hence, the simulation of groundwater resources is necessary to (a) determine the quantity available for water supply and (b) estimate probable nitrate contamination to propose subsequent remediation techniques to protect public health. Furthermore, the aquifer's heterogeneity as well as the lack of hydraulic conductivity data, in a large-scale study area, creates uncertainty regarding groundwater flow and nitrate transport simulation. Deterministic modelling approaches for spatially distributed nitrate concentration simulation could not estimate the contamination risk, since it can address only one realization of the aquifer. This study estimates the effect of hydraulic conductivity uncertainty on the simulation of groundwater nitrate concentration. The proposed framework uses Geostatistical Sequential Gaussian Simulation (SGSIM) for the generation of equally probable realizations of the spatial distribution of hydraulic conductivity. It includes the application of a modelling system based on the following inter-linked models: a rainfall-runoff model, a reservoir operation model, a lake-aquifer interaction model, a groundwater flow model, and a nitrate transport and dispersion model. The last two models simulate the multiple realizations of aquifer's groundwater flow and nitrate concentration. Furthermore, two analyses (a statistical and a threshold analysis) are employed to estimate the exceedance probability of the nitrate concentrations and their spatial extent. The results indicate that hydraulic conductivity uncertainty does affect the simulation of nitrate concentration and that nitrate concentrations will most probably exceed the thresholds in areas where groundwater is extracted for domestic use.

Keywords: Geostatistical tools; Greece; Hydraulic conductivity uncertainty; Nitrate pollution; Statistical and threshold analysis; Stochastic simulation.

MeSH terms

  • Environmental Monitoring
  • Greece
  • Groundwater*
  • Nitrates / analysis
  • Uncertainty
  • Water Pollutants, Chemical* / analysis

Substances

  • Nitrates
  • Water Pollutants, Chemical