For reconstruction of forest cover change in the Polish Carpathians we plan to use the 2nd Austro-Hungarian topographic military survey carried out around 1850 (scale of 1:28800). For the 1930s, Polish military maps (1:100000) are available. For the second half of the 20th century we will use information from modern Polish topographic maps (1:10000 to 1:100000). For the Swiss Alps we will use the so-called Dufour Map at scales of 1:25000 for the lowlands and 1:50000 for the Alpine areas, started in the 1840s; the Siegfried Maps (1:25000 – lowlands, 1:50000 – Alpine areas) started in the 1880s and updated in regular intervals until 1940s; and finally modern Swiss topographic maps (Landeskarte, 1:25000) started in 1950 and updated continuously. The two latter maps have been frequently used for land cover change studies. Application of the Dufour Map has been restricted to comparative approaches so far. We will also test approaches to automatically extract forest / non-forest (F/NF) information from topographic maps using image morphological analysis. A wall-to-wall mapping approach will be applied.
For the period since mid-20th century we will complement information from maps with aerial and high resolution satellite imagery. Remote sensing data will be used also to identify different succession stages from open land to closed forest. For selected areas with detailed land use history information we will match the different succession stages with time since abandonment. The spatial sampling strategy will be applied. Stratified samples of test areas (administrative units for which consistent temporal sets of census data are accessible) representing various environmental, socio-economic conditions and land use histories will allow to test hypotheses on the importance of specific factors to forest cover dynamics.
ALS data allow to automatically estimate the height of various terrain objects including vegetation with a high accuracy. In this way, ALS data may help to identify abandoned areas at various successional stages without a tedious and error-prone visual interpretation of aerial imagery. Due to high cost of ALS data, this task will be carried out for selected sites within case study areas defined previously in task 4.
Data on past climate change in Swiss Alps & Polish Carpathians will be based on existing studies (meta-analysis) for the 19th century, while climate maps will be developed at fine spatial resolution for the 20th and 21st century via  spatial extrapolation of long-term monthly mean climate-terrain relationships from existing climate stations, and  superimposition of monthly time series to long-term means by scaling monthly climate anomalies from coarser (CRU TS 1.2, ENSEMBLES regional circulation model outputs) to finer resolution of current climate means. To disentangle the effect of climate change from the effects of land use change on forest cover change in the past case study areas studied in tasks 3-5 will be compared, in particular those located at the timberline. Therefore, we will calibrate modern statistical regression models, where forest cover change is used as a dependent variable, and climate and land use change information are used as predictors to explain forest cover change.
Existing models for land use projections (Switzerland) are based on trends extracted from federal land use statistics (1980s / 1990s). The models distinguish between 3 land-use scenarios: business as usual (continuation of observed trends between land use statistics), liberalization (no more public support for agricultural production and conservationorientated agriculture), and extensification (increased support for conservation-oriented agriculture). Models will be refined with new long-term forest cover change perspective and knowledge of the climate / land use contributions to forest cover change, and then run for various climate and land-use scenarios.
Applying a model developed in Swiss Alps in the Carpathians requires analysis of ancillary data availability, comparability of model foundations and extensive testing. We plan to collect necessary ancillary data (e.g., socio-economic) and verify the model in Polish Carpathians using the past forest cover data, and then, under the assumption of model validity, develop forest cover change estimates for Polish Carpathians.
Carbon pools will be quantified using IPCC methodology applying region-specific values for typical tree species composition, stand densities and biomass expansion factors, for Swiss Alps and Polish Carpathians, using past forest cover trends and future predictions. Impacts for biodiversity will be assessed via calculation of forest pattern indices developed with morphological analysis of forest maps, and analysis of their changes over time.
Project supported by a grant from Switzerland through the Swiss Contribution to the enlarged European Union.
Value of co-financing: 681 420, 22 CHF.