Water balance and nitrate leaching from arable land in a changed climate : a model study

Abstract: This thesis aims to present the essential background on how to perform climate changeimpact assessments, and to present the results from a climate impact assessment on waterbalance and nitrate leaching for an arable Swedish soil. The soil is a sandy soil in southwesternSweden, grown with spring cereals. This study is meant to be a benchmark example,and cannot be seen as a regional or national assessment for Sweden, rather as an approachto present and analyze the most important parts of these kinds of assessments.A dynamical simulation model (COUP, Jansson and Karlberg, 2004) was used for thisstudy. The model was parameterized and calibrated against data from an experimental site,located in Mellby in Hallands county, south western Sweden. Measurements were carriedout between 1st of April 1988 and 1st of April 1991. The data set consists of daily standardweather data and discharge, data on soil water content, soil nitrogen and nitrogen contentsin drainage water from to experimental fields grown with spring cereals. The model wascalibrated against the 4-year data set based on a GLUE-procedure in which a number of“acceptable parameter sets” were identified. One of these parameter sets were randomlychosen for the climate impact model runs performed in this study. The driving data for themodel are 30-year climate data, including data for precipitation, temperature, solar radiation,relative humidity and wind speed, which enables long-term simulations of water andnitrogen flows. Three different simulations were performed, one for present climate as areference scenario with climate data from 1971-2000, and two different emission scenariosrepresenting year 2071-2100. The driving data were constructed by the delta-change method,which is a common way of interfacing regional climate model output with impactmodels.Results show that, for both scenarios, that nitrate leaching will increase with 41 % and66 % respectively. This is mainly due to increased winter temperatures (increasing nitrogenmineralization with 22 % and 32 % respectively) and increased drainage (20 % and 33% respectively) during the period when the soil is left bare.It is important to remember that the study includes many generalizations, both in parameterizationand in driving data. Despite that, the approach with a dynamical model drivenwith long-term climate data is a very robust and valuable way of making such assessments.Further studies need to consider crop growth characteristics and crop parameterization tobe able to simulate growth of other varieties more suitable in a changed climate. Ensemblemodeling can also be an approach to reduce biases in the modeling chain.