Estimation of past forest cover changes over last 150 years: Case studies from the Swiss Alps and the Polish Carpathians

Transcription

1 Estimation of past forest cover changes over last 15 years: Case studies from the Swiss Alps and the Polish Carpathians Katarzyna Ostapowicz 1, Monika Dobosz 1, Dominik Kaim 1, Jacek Kozak 1, Krzysztof Ostafin 1, Agnieszka Wypych 2, Zbigniew Ustrnul 2, Urs Gimmi 3, Matthias Bürgi 3, Achilleas Psomas 3, Niklaus Zimmerman 3 1 Department of GIS, Cartography and Remote Sensing, Institute of Geography and Spatial Management, Jagiellonian University, Kraków, Poland 2 Department of Climatology, Institute of Geography and Spatial Management, Jagiellonian University, Kraków, Poland 3 Swiss Federal Research Institute WSL, Zuercherstrasse 111, 893 Birmensdorf, Switzerland PROJECT SUPPORTED BY A GRANT FROM SWITZERLAND THROUGH THE SWISS CONTRIBUTION TO THE ENLARGED EUROPEAN UNION

2 Background over the past decades agricultural land in Europe has declined and forest area has expanded considerably recent trends of land abandonment have been most pronounced in marginal areas like mountains (agriculture and economical inefficient) the potential upper tree line was shifting upward as a consequence of global warming (e.g. the Alps) studies conducted so far on forest cover change in the Swiss Alps and the Polish Carpathians were either (1) local case studies restricted in their spatial extent, (2) based on forest statistics and not spatially explicit, or (3) looked only at recent changes spatially accurate large-scale and long-term reconstructions of forest cover changes are lacking for both mountain regions and the extent, spatial or temporal variation and underlying driving forces of the forest cover change trends are still insufficiently understood

3 Project

4 Study areas Study Area

5 Study area PL CZK PL UKR PL CH PL AUT PL PL I SLK PL Polish Carpathians AREA: 2 sqkm Alt range 2-25 Swiss Alps AREA: 1 sqkm Alt range: 25-4

6 Aim The aim of this part of project is to assess influence of different drivers like land use and/or climate change on forest cover changes over the last 15 years in the parts of two European mountain ranges: the Swiss Alps and the Polish Carpathians early 2th century early 21st century

7 Database: forest cover Data (the Polish Carpathians) o o o Second Austro-Hungarian military survey maps (~185, scale 1:28 8) Polish topographic maps (193s, 1:1 ) Polish topographic maps (197s, 1: 25 ) o Landsat satellite data (211, spatial resolution 3 m) Data (the Swiss Alps) o Dufour Map (~185, scale 1:25, 1:5,) o Siegfried Map (edition 188 and 194, scale 1:25, 1:5,) o Landeskarte der Schweiz (197s, 1: 25,) o Landeskarte der Schweiz (212, 1: 25,)

8 FOREST PL CH forest cover 186s forest cover 185s forest cover 193s forest cover 188s forest cover 197s forest cover 194s forest cover forst cover 211 forest cover 21 forest cover change 186s- 193s forest cover change 185s- 188s forest cover change 193s- 197s forest cover change 188s- 194s forest cover change forest cover change 197s- 211 forest cover change 194s- 21 Database: factors/drivers context topography & climate FACTORS/DRIVERS source (PL) environmental factors topography altitude SRTM DEM (9 m) slope norhtness eastness climate Mean values for temperature (185s-21(11)) CARPAT CLIM Mean values for precipitation (185s-21(11)) Mean annual DDsum (185s- 21(11)) socio-economic population number of people statistic context agriculture economy accessibility forest composition and configuration number of housholds age (different categories) number of farms number of animals (different types: cows,pigs, sheep, goats etc) farming area (cropland/pasture) number of eployees by sector railroad (yes/no)/main road (yes/no) distance to settelments forest spatial pattern distance to forest al books (for analysis points of time) source (CH) DHM1 WORLD CLIM, CRU, DAYMET statistic al data (for analysis points of time) forest forest maps maps (for (for analysis analysis points points of time) of time) socioeconomic

9 Methods units and models Communities: 199 Districts: 15 Cantons: 5 Carpathians communes (CC list): 234 cities and villagies about 1862: 284 cities and villagies 28: 248 statistical models (General Linear Models; GLMs)

10 Results

11 Results persistent loss increase (ha)

12 Results portion of total landscape [%] brzozowski strzyżowski tarnowski myślenicki nowotarski żywiecki

13 Results NW OW URI GL GR portion of total landscape,4,35,3,25,2,15,1,5, GR GL URI OW NW tot forest cover 185 forest cover 188 forest cover 194 forest cover 212

14 Results (the Polish Carpathians): Forest cover change communes and R 2 single models min. elevation [m] change of forest proportion slope degree [o] northness eastness distance to forest [m] 186s-193s,23,33,22,22,25 193s-197s,24,2,2,3,25 197s-211,24,13,5,25 net forest increase 186s-193s,25,64,2,24,25 193s-197s,25,26,9,15,25 197s-211,25,24,16,25 186s-193s,24,47,19,23,25 193s-197s,24,11,6,1,25 197s-211,24,5,4,1,25 net forest decrease 186s-193s,8,25,2,21,25 193s-197s,4,3,25 197s-211,11,59,7,19,25 forest decrease [%] 186s-193s,8,5,2,25 193s-197s,11,31,9,13,25 197s-211,3,25,1,9,25 whole model change of forest proportion 186s-193s,3 193s-197s,2 197s-211,14 net forest increase 186s-193s,66 193s-197s,26 197s-211,59 186s-193s,52 193s-197s,14 197s-211,28 net forest decrease 186s-193s,29 193s-197s,5 197s-211,32 forest decrease [%] 186s-193s,42 193s-197s,44 197s-211,9

15 Results (the Polish Carpathians): Forest increase and slope degree 186s-193s 193s-197s 197s-211 net forest increase 1,9,8,7,6,5,4,3,2,1 1 2 net forest increase 1,9,8,7,6,5,4,3,2,1 1 2 net forest increase 1,9,8,7,6,5,4,3,2,

16 Results (the Polish Carpathians): Forest decrease and slope degree 186s-193s 193s-197s 197s net forest increase,8,6,4,2 net forest increase,8,6,4,2 net forest increase,8,6,4,

17 Results (the Polish Carpathians): Change in forest proportion and elevation 4 186s-193s min. elevation [m] s-197s min. elevation [m] s min. elevation [m]

18 R² =, min. elevation [m] R² =, min. elevation [m] Results (the Polish Carpathians): forest cover and elevation R² =, min. elevation [m] R² =, min. elevation [m]

19 Results (the Swiss Alps): Forest increase Model: Target variable: Explanatory variables: GLM (binomial) stepwise, sample: 1 nonforest pixels at t1 forest gain (yes/no) exposition (northeness/eastness), altitude, slope, distance to forest edge at 1st time step, distance to settlement,4,35,3 Adj D2,25,2,15,1,5 OM-sieg1st ( ) sieg1st-sieglast ( ) sieglast_today (194-21)

20 Results (the Swiss Alps): Forest decrease Model: Target variable: Explanatory variables: GLM (binomial) stepwise, sample 1 forest pixels at t1 forest gain (yes/no) exposition (northeness/eastness), altitude, slope, distance to forest edge at 1st time step, distance to settlement,4,35 Adj D2,3,25,2,15,1,5 OM-sieg1st ( ) sieg1st-sieglast ( ) sieglast_today (194-21)

21 Results (the Swiss Alps): forest cover Model: Target variable: Explanatory variables: GLM (binomial) stepwise, sample 1 of all pixels forest (yes/no) exposition (northeness/eastness), altitude, slope, forest cover at previous time step,8,7,6,5 Adj D2,4,3,2,1 topo only incl previous fcover OM (185) sieg1st (188) sieglast (194) today (21)

22 Conclussions net forest cover increase have been observed for both study areas and across almost all periods long-term studies help to identify hot-spots of change and persistency (spatial) and periods of change and stability (temporal) forest-transition -> different stories at regional and locale scales relatively, topography and context information do not significantly contribute to the explanation of forest cover decrease or increase

23 Thank you for your attention!