Comparison of Polish and German maps of agricultural soil quality using GIS

Abstract Soil maps serve different purposes in rural planning, environmental protection, and agricultural production. On the scale of the European Union, soil‐quality maps are used as main input‐data source in the process of “less favored area's” (LFA) delineation, which in course get special financial support. However, every EU member state uses its own national soil‐map resources to derive soil‐quality estimates, as no uniform mapping scheme or map resources exist for the whole territory. Information about the actual comparability of these map resources is missing. Because of historical reasons, for some parts of the Republic of Poland soil‐quality maps exist, which were edited using the German mapping scheme. In this study, we used nine German map sheets at a generalized scale of 1:25,000 and compared the German soil‐map attributes (German: “Bodenschätzung”) with the respective Polish information (Polish: “mapa glebowo‐rolnicza”) at the same scale. Map analysis of 94,000 ha in a study area near Płock, Central Poland, was performed on vector data that was digitized from original raster maps in a Geographic Information System (GIS). An automatically produced, stratified‐random point‐sampling scheme was used to statistically compare the map attributes. The distance of each sampling point to the vector polygon borders of the two maps in the GIS served for continuously excluding data points from the analysis to acknowledge spatial uncertainties due to mapping accuracy and other factors. Qualitative soil‐texture information was translated into water‐storage‐capacity values. Correlation coefficients between Polish and German arable‐soil‐quality assessments were between 0.57 for all data points and 0.72, when the data set was restricted to points with minimum distance of 60 m to any polygon border. Further increase of spatial restrictions did not improve correlation coefficients. Correlations between soil‐quality assessments and water‐storage capacities of the respective other national mapping scheme fell between 0.53 and 0.71. Higher amounts of variation of German soil quality could be explained using a factorial model with soil‐texture and soil‐type information, than by simple linear regression of quality assessments. Soil water‐storage capacities for German soil‐texture information resulted in average estimates, which were only 70%–88% of the respective Polish values, but only 41% in the case of sandy soil texture. Our analysis indicated that information of the Polish soil‐quality map with regard to soil texture in different depths could contribute to a more refined and differentiated quantitative assessment of agricultural soil quality.