Uncertainty in simulated nitrate leaching due to uncertainty in input data. A case study

Abstract

Deterministic leaching models are used to estimate regional losses of nitrate from agricultural land to the environment. The estimated leaching losses are associated with uncertainty arising from uncertainty in the input data used. In the present case study we have assessed this uncertainty by use of Monte Carlo analysis, using the Latin hypercube sampling technique. Input data have preferably been adopted from publicly available data. Data which could not be retrieved from the databases was assessed by guided estimates or based on local data.The estimated annual leaching loss from the study region was around 106 kg N ha 71 , which is in agreement with previous findings. The uncertainty in the leaching expressed in terms of coefficients of variation (CV) depended on the agricultural practices. CV's for arable farm rotations, cattle farm rotations, and pig farm rotations were around 20, 30 and 40%, respectively. Breakdown of the total uncertainty into contributions of different error sources did not isolate one single all important source.

Uncertainty in Simulation of Nitrate Leaching at Field and Catchment Scale within the Odense River Basin

“…This imposes limitations as to the interpretation of water flow and leaching processes at spatial scales smaller than the catchment scale—that is, field level—but is not expected to severely impact the lumped discharge response at the monitoring station used in this study. Hansen et al (1999) varied precipitation in the Karup Basin by adding a random error (zero mean and a standard deviation of 50%) to measured precipitation series and found that precipitation contributed least to the uncertainty associated to water balance, mineralization, and plant N uptake. Thorsen et al (2001) for the same Karup Basin introduced the same uncertainty to the precipitation time series as in Hansen et al (1999) and found that this affected the simulated yearly water balance but not the simulated nitrogen balance, suggesting that the timing of the percolated water controlled by the soil hydraulic properties is more important for the simulated nitrogen loads than the total annual amounts of percolation.…”

“…Hansen et al (1999) performed an uncertainty analysis for nitrate leaching from the rootzone using the DAISY model (Hansen et al, 1991) for several land management types in the Karup Basin in the western part of Denmark. Thorsen et al (2001) followed up on the studies by Refsgaard et al (1999) and Hansen et al (1999) on catchment‐scale modeling of nitrate leaching for the Karup Basin, applying the MIKE‐SHE/DAISY (Styczen and Storm, 1993) code to the Karup Basin using Monte Carlo techniques for assessing model input parameter uncertainty for the same five key parameters as in Hansen et al (1999) These parameters—precipitation, soil hydraulic properties, soil organic matter content, slurry composition and depth of the oxidized zone (redox interface) in the aquifer—are known to be important for water balance, nitrate leaching, and transformation. Thorsen et al (2001) concluded that uncertainty in simulated nitrate flux concentrations depends on the considered spatial and temporal scale; that is, uncertainty in simulated flux concentrations from the rootzone is much larger than simulated concentrations at the catchment scale.…”

“…Christiaens and Feyen (2001, 2000) used LHS for an uncertainty analysis on hydraulic properties and propagation through the distributed MIKE‐SHE model (Refsgaard and Storm, 1995). As mentioned above, Hansen et al (1999) and Thorsen et al (2001) also used LHS for parameter distribution propagation in the MIKE‐SHE/DAISY model.…”

Soil and informatics science combine to develop S-map: A new generation soil information system for New Zealand

“…Smap will be capable of providing alternative spatial realisations of soil properties of interest that can be used in stochastic uncertainty analysis modelling, e.g. (Hansen et al, 1999;Heuvelink et al, 2010) …”

Assessing Uncertainty Propagation through Physically Based Models of Soil Water Flow and Solute Transport

“…The Monte Carlo method has been widely used to propagate uncertainties through soil hydrological models and has become increasingly popular over recent years as the computational overheads of performing such an analysis have declined. Examples in soil hydrology can be found in Petach et al (1991); Bennett et al (1998); Duke et al (1998); Hansen et al (1999); Kros et al (1999); Dillah and Protopapas (2000); Thorsen et al (2001); Keller et al (2002); De Vries et al (2003). For distributed modeling of soil water flows and solute transport, the sampling algorithm must accommodate spatial and temporal dependencies in model inputs and parameters (De Roo et al, 1992;Endreny and Wood, 2001).…”