Multispectral remotely sensed data in modelling the annual variability of nitrate concentrations in the leachate
Introduction
Agricultural activities have been identified as major sources of non-point source pollutants of ground and surface waters (Clark et al., 2004, Galloway et al., 2002, Kunkel et al., 2005). Especially nitrogen is one of the most problematic nutrients (Basnyat et al., 2000, Perry and Vanderklein, 1996). With the Water Framework Directive (Directive 2000/60/EC), the Nitrates Directive (Directive 91/676/EEC) and further activities the European Union has implemented procedures aiming at recovering a good quality of water resources in 2015 (Beaudoin et al., 2005, Letcher and Giupponi, 2005). To describe the status quo, to analyse the interdependencies between agriculture and hydrosphere and to investigate the effects of agricultural-environmental reduction measures, the use of mathematical models is a common approach (Galloway et al., 2002, Meynard et al., 2002). In general, for integrated analyses often a coupling of specialised models and the development of interdisciplinary model systems is required (Roetter et al., 2007, Lehtonen et al., 2007). But due to different reference parameters, scales, model philosophies, etc., a direct combination of models as system components is not possible without any difficulty. In this context, the accurate development and further enhancement of model interfaces is of utmost importance (Ahrends et al., 2008).
For the analysis of diffuse nitrogen pollution, a number of models and model systems with different intentions and complexity have been developed during the last decades as described by Behrendt et al., 2000, Birkinshaw, 2000, Matejicek et al. (2003) or Wriedt and Rode (2006). During the project Management of Regional German River Catchments (REGFLUD), which was funded by the German Federal Ministry of Education and Research (BMBF), a combined agricultural-economic and hydro(geo)logic model system has been developed (Goemann et al., 2005, Kunkel et al., 2005, Wendland et al., 2005). This model system, henceforth named REGFLUD, is composed of the standalone models RAUMIS (Henrichsmeyer et al., 1996) and GROWA (Kunkel and Wendland, 2002). The interface (Wendland et al., 2005) between the agricultural-economic model RAUMIS and the hydrological model GROWA, which is based on nitrogen surpluses, was developed on the basis of CORINE Land Cover (Bossard et al., 2000). For the combination and the geographical referencing in a GIS database it has to be taken into account that the models display different regional resolutions—raster cells in the hydrological model and administrative units in the agricultural-economic model (Wendland et al., 2005).
As the land use has been found to have a large effect on the amount of nitrogen exported to the hydrosphere (Buck et al., 2004, Jordan et al., 1997, Poor and McDonnell, 2007, Salvia-Castellvi et al., 2005), multispectral remotely sensed data have been used to enhance the knowledge of land use and thereby improve the results of nitrogen models (Davenport et al., 2003, Jongschaap, 2006, Mattikalli and Richards, 1996, Pandey et al., 2005, Payraudeau et al., 2004). In this study satellite imagery from the sensors ASTER, SPOT and LANDSAT have been used in order to map the land cover, in particular different agricultural crops, specifically, to identify crop rotation over the period of 2000–2004.
Three principal types of satellite imagery application are conceivable to enhance the interdisciplinary REGFLUD model system. The first is a substitution of input parameters, the second is an interface between two stand-alone models from different scientific disciplines, and the third is a calibration and extension of an existing individual model by using remotely sensed data. In the present work, satellite imagery has been used in order to (i) substitute the low resolution and inadequately differentiated CORINE data by high resolution land cover and imperviousness maps, (ii) perform a crop-specific differentiation of the district-based RAUMIS nitrogen surpluses, and (iii) calibrate and extend coefficients for real evapotranspiration calculations within GROWA. This results in an enhancement of a GIS-based estimation of nitrate concentration in the leachate verified by a comparison of modelled nitrate concentration in the leachate with observed nitrate concentrations at the groundwater surface.
Section snippets
The study area
Water management on the European Level is based on river basins. To analyse the potential of multispectral satellite imagery in interdisciplinary model systems, the catchment basin of the Rur River, a subcatchment basin of the River Maas (Meuse) located at the Belgian–Dutch–German border near the town of Aachen was chosen (Fig. 1). The Rur River, with its length of 165 km, drains a total area of 2354 km2. The catchment basin can be separated into two main regions: The Southern part covers the
The REGFLUD model system—general approach
The REGFLUD model system (Goemann et al., 2005, Kunkel et al., 2005, Wendland et al., 2005) combines models from different scientific disciplines. Fig. 2 shows the integration of the agricultural-economic model RAUMIS with the hydrological models GROWA and WEKU. RAUMIS (Henrichsmeyer et al., 1996) uses a set of agro-environmental indicators to represent the agricultural sector, taking into account agricultural statistics of administrative districts. Indicators such as fertiliser surplus
Processing and analysis of remotely sensed data
The first enhancement in the interdisciplinary model system is the substitution of required input parameters by data sets with a higher spatial and content-related resolution. Therefore, data recorded by the sensors SPOT 1/2 have been acquired for the years 2000, 2002 and 2003. In addition, imagery from LANDSAT 7 ETM+ for 2001 and from ASTER for 2004 has been procured in order to obtain cloudless imagery during the relevant phenological phase (May to August). In order to ingest the remote
Disaggregation of nitrogen surpluses given by RAUMIS on remotely-sensed agricultural crops
Building on these crop-specific land cover maps, the next level of model enhancement can be performed. Since nitrogen surpluses govern the coupling of RAUMIS and GROWA, an enhancement of the nitrogen surplus database is mandatory. As mentioned before, the agricultural nitrogen surpluses offered by RAUMIS were previously transferred to the CORINE Land Cover map (Wendland et al., 2005). This means, for example, that the average nitrogen surplus of 50 kg ha−1 yr−1 provided by RAUMIS for the district
Assimilation of a remotely-sensed agricultural crop map for calibration and extension of the GROWA model
A third enhancement of the REGFLUD model system involves a modification of one of its system components. A large improvement potential is seen in the way in which ETr is calculated in GROWA. Promising approaches for estimating ETr with a feasible implementation into GROWA can be found in Bastiaanssen et al., 1998, Boegh et al., 2002, Carlson et al., 1995, Kite and Droogers, 2000, Nagler et al., 2005 or Nishida et al. (2003). Remotely sensed data together with classical micrometeorological flux
Enhanced REGFLUD model system using remotely sensed data
The aim of this study was to enhance and extend the REGFLUD model system by using remotely sensed data. This enhancement is demonstrated in Fig. 6. The source of the enhancements can be found in the linkage data set nitrogen surpluses, in the GROWA model results and finally in the output of the model system, the nitrate concentration in the leachate.
After calculating the denitrification in soil with DENUZ using the enhanced GROWA model outputs as well as disaggregated nitrogen surpluses, the
Scenario analysis
To locate the REGFLUD-model system in the context of policy advice and to evaluate the effects of political nitrogen reduction measures, two scenarios were developed and analysed. In scenario 1 nitrogen reduction is encouraged by raising a tax of 200% on mineral nitrogen fertilisers. In scenario 2 a limitation of the cattle stocking rate to 1 livestock unit per hectare is imposed (Wendland et al., 2005). The effects of these agro-environmental policy options were analysed with and without the
Conclusion and outlook
The REGFLUD model system has been constructed in order to develop and apply multi-criteria scientific methods which are able to predict diffuse pollution in river basins taking into account economic feasibility and social acceptability. Multispectral remotely sensed data have been integrated into the model system to enhance the input data sets land cover and impervious surfaces, to enhance the model interface nitrogen surplus and to enhance the system component GROWA. Through the improvement of
References (56)
- et al.
Coupled hydrological-economic modelling for optimised irrigated cultivation in a semi-arid catchment of West Africa
Environmental Modelling & Software
(2008) - et al.
The use of remote sensing and GIS in watershed level analyses of non-point source pollution problems
Forest Ecology and Management
(2000) - et al.
A remote sensing surface energy balance algorithm for land (SEBAL)—1. Formulation
Journal of Hydrology
(1998) - et al.
Nitrate leaching in intensive agriculture in Northern France: Effect of farming practices, soils and crop rotations
Agriculture Ecosystems & Environment
(2005) - et al.
Evaluating evapotranspiration rates and surface conditions using Landsat TM to estimate atmospheric resistance and surface resistance
Remote Sensing of Environment
(2002) Modelling nitrate transport in the Slapton Wood catchment using SHETRAN
Journal of Hydrology
(2000)- et al.
Distributed modeling of groundwater recharge at the macroscale
Ecological Modeling
(2005) - et al.
Scale-dependence of land use effects on water quality of streams in agricultural catchments
Environmental Pollution
(2004) - et al.
The use of earth observation techniques to improve catchment-scale pollution predictions
Phys. Chem. Earth
(2003) Run-time calibration of simulation models by integrating remote sensing estimates of leaf area index and canopy nitrogen
European Journal of Agronomy
(2006)