Vědci z JČU se zúčastnili dvou stáží ve dnech 13.-24.8. 2012 a 17.6.-25.7. 2013 na Institutu Environmentálních studií v Amsterdamu: zpracování prostorových dat a simulace změn ve využívání území pro účely disertace – první část zahraniční stáže v Amsterdamu.

Metoda zpracování prostorových dat je shrnuta v konferenčním článku "DRIVERS OF AGRICULTURAL ABANDONMENT: A CASE STUDY FROM THE ŠUMAVA MOUNTAINES, CZECH REPUBLIC" Jitky Strakové:

Contour lines from ZABAGED were used to calculate altitude, slope and aspects in ArcGIS 10.0. The resulting raster datasets were sized at 20x20 meters pixels to cover the smallest land plots. Raster data on aspects were reclassified into three categories referring to the south, southeast and southwest orientation followed by orientation to the west and east, while the last category gathered the pixels oriented to the north, northwest and northeast. Within ArcGIS 10.0, distance to roads, pathways and human settlement were measured by Euclidean Distance. Information on ownership status was manually attributed from the public database of owners managed by the Czech Bureau of Land Survey and Register and reclassified into categories of private owners, state owners and NP and PLA Šumava. Data on NP and PLA zonation were maintained from the Administration of NP and PLA Šumava and were spatially referenced in ArcGIS 10.0.

The selection of potential biophysical drivers of agricultural abandonment was forced by the literature review, but also it became a subject to data availability. Once all the GIS maps including the land use, land tenure, nature protection zonation, elevation, slope, aspects and distances to households and farmsteads, roads and paths were completed, the individual maps were converted firstly into raster format with a cell size of 20 m and secondly into ASCII format in ArcGIS 10.0.. Balanced sample of all biophysical explanatory variables was chosen and introduced in the statistical model to analyse which variables contribute significantly to the explanation of land abandonment. Following the logic of land use (occurrence of either agriculturally used or abandoned grassland), simple logistic approach is used. Statistics were calculated in SPSS for Windows. The ROC (relative operating characteristic or receiver operating characteristic) measure evaluates the performance of model. ROC value higher than 0.7 is considered acceptable for LUCC modelling (Leschen et al ,2005). Pearson’s test was used to estimate inter-correlations between the independent variables. A value of 0.8 or greater is considered to be a measure of high correlation that demands exclusion of one from the pair of the interlinked variables.

Druhá část zahraniční stáže v Amsterdamu byla na účel zpracování scénářových studií změn ve využívání území pro účely disertace a příprava struktury článku.

Na stáži byl vytvořen základ pro 4 scénářové studie: - Adaptation to new CAP - National Park is closed down - EU is broken up - Continuation of present trend