Spatial aggregation for crop modelling at regional scales: the effects of soil variability

Abstract

Modelling agriculture production and adaptation to the environment at regional or global scale receives much interest in the context of climate change (CC). One concern is to take into account the spatial variability of the environmental conditions (e.g. climate, soils, management practices) used as model input because the impacts of CC on cropping systems depend strongly on the site conditions [1]. For example CC effects on yield can be either negative or positive depending on the soil type [2]. Additionally, the use of different methods of upscaling and downscaling adds new sources of modelling uncertainties [3].In the present study, the effect of aggregating soil data by area majority of soil mapping units was explored for regional simulations with the soil-vegetation model CoupModel for a region inGermany (North Rhine-Westphalia). Data aggregation effects (DAE) were analysed for wheat yield, water drainage, soil carbon mineralisation and nitrogen leaching below the root zone. DAE were higher for soil C and N variables than for yield and drainage and were strongly related to the presence of specific soils within the study region. These 'key soils' were identified by a model sensitivity analysis to soils present in the region. The spatial aggregation of the key soils additionally influenced the DAE. A spatial analysis of the pattern of these key soils (i.e. presence/ absence, coverage and aggregation) can help in defining the appropriate grid-resolution that would minimize the error caused by aggregated soil input data in regional model simulations. In a second step the method will be applied and evaluated with respect to another European region(Tuscany) which is characterised by a warmer and drier climate.[1] Kersebaum, K.C., Nendel, C., 2014. Site-specific impacts of climate change on wheat production across regions ofGermany using different CO2 response functions. Eur. J. Agron. 52, 22–32. doi:10.1016/j.eja.2013.04.005[2] Folberth, C. et al, 2016. Uncertainty in soil data can outweigh climate impact signals in global crop yield simulations. Nat. Commun. 7, 11872. doi:10.1038/ncomms11872[3] Ewert et al., 2011. Scale changes and model linking methods for integrated assessment of agri-environmental systems. Agric. Ecosyst. Environ. 142, 6–17. doi:10.1016/j.agee.2011.05.016