@misc{DallmeyerClaussenNietal.2017,
  author    = {Dallmeyer, Anne and Claussen, Martin and Ni, Jian and Cao, Xianyong and Wang, Yongbo and Fischer, Nils and Pfeiffer, Madlene and Jin, Liya and Khon, Vyacheslav and Wagner, Sebastian and Haberkorn, Kerstin and Herzschuh, Ulrike},
  title     = {Biome changes in Asia since the mid-Holocene},
  series = {Postprints der Universit{\"a}t Potsdam : Mathematisch Naturwissenschaftliche Reihe},
  journal   = {Postprints der Universit{\"a}t Potsdam : Mathematisch Naturwissenschaftliche Reihe},
  number    = {643},
  issn      = {1866-8372},
  doi       = {10.25932/publishup-41875},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-418755},
  pages     = {107 -- 134},
  year      = {2017},
  abstract  = {The large variety of atmospheric circulation systems affecting the eastern Asian climate is reflected by the complex Asian vegetation distribution. Particularly in the transition zones of these circulation systems, vegetation is supposed to be very sensitive to climate change. Since proxy records are scarce, hitherto a mechanistic understanding of the past spatio-temporal climate-vegetation relationship is lacking. To assess the Holocene vegetation change and to obtain an ensemble of potential mid-Holocene biome distributions for eastern Asia, we forced the diagnostic biome model BIOME4 with climate anomalies of different transient Holocene climate simulations performed in coupled atmosphere-ocean(-vegetation) models. The simulated biome changes are compared with pollen-based biome records for different key regions. In all simulations, substantial biome shifts during the last 6000 years are confined to the high northern latitudes and the monsoon-westerly wind transition zone, but the temporal evolution and amplitude of change strongly depend on the climate forcing. Large parts of the southern tundra are replaced by taiga during the mid-Holocene due to a warmer growing season and the boreal treeline in northern Asia is shifted northward by approx. 4 degrees in the ensemble mean, ranging from 1.5 to 6 degrees in the individual simulations, respectively. This simulated treeline shift is in agreement with pollen-based reconstructions from northern Siberia. The desert fraction in the transition zone is reduced by 21\% during the mid-Holocene compared to pre-industrial due to enhanced precipitation. The desert-steppe margin is shifted westward by 5 degrees (1-9 degrees in the individual simulations). The forest biomes are expanded north-westward by 2 degrees, ranging from 0 to 4 degrees in the single simulations. These results corroborate pollen-based reconstructions indicating an extended forest area in north-central China during the mid-Holocene. According to the model, the forest-to-non-forest and steppe-to-desert changes in the climate transition zones are spatially not uniform and not linear since the mid-Holocene.},
  language  = {en}
}
@article{BartholdWiesmeierBreueretal.2013,
  author    = {Barthold, Frauke Katrin and Wiesmeier, Martin and Breuer, L. and Frede, Hans-Georg and Wu, J. and Blank, F. Benjamin},
  title     = {Land use and climate control the spatial distribution of soil types in the grasslands of Inner Mongolia},
  series = {Journal of arid environments},
  volume    = {88},
  journal   = {Journal of arid environments},
  number    = {1},
  publisher = {Elsevier},
  address   = {London},
  issn      = {0140-1963},
  doi       = {10.1016/j.jaridenv.2012.08.004},
  pages     = {194 -- 205},
  year      = {2013},
  abstract  = {The spatial distribution of soil types is controlled by a set of environmental factors such as climate, organisms, parent material and topography as well as time and space. A change of these factors will lead to a change in the spatial distribution of soil types. In this study, we use a digital soil mapping approach to improve our knowledge about major soil type distributing factors in the steppe regions of Inner Mongolia (China) which currently undergo tremendous environmental change, e.g. climate and land use change. We use Random Forests in an effort to map Reference Soil Groups according to the World Reference Base for Soil Resources (WRB) in the Xilin River catchment. We benefit from the superior prediction capabilities of RF and additional interpretive results in order to identify the major environmental factors that control spatial patterns of soil types. The nine WRB soil groups that were identified and spatially predicted for the study area are Arenosol, Calcisol, Cambisol, Chernozem, Cryosol, Gleysol, Kastanozem, Phaeozem and Regosol. Model and prediction performances of the RF model are high with an Out-of-Bag error of 51.6\% for the model and a misclassification error for the predicted map of 28.9\%. The main controlling factors of soil type distribution are land use, a set of topographic variables, geology and climate. However, land use and climate are of major importance and topography and geology are of minor importance. The visualizations of the predictions, the variable importance measures as result of RF and the comparisons of these with the spatial distribution of the environmental factors delivered additional, quantitative information of these controlling factors and revealed that intensively grazed areas are subjected to soil degradation. However, most of the area is still governed by natural soil forming processes which are driven by climate, topography and geology. Most importantly though, our study revealed that a shift towards warmer temperatures and lower precipitation regimes will lead to a change of the spatial distribution of RSGs towards steppe soils that store less carbon, i.e. a decrease of spatial extent of Phaeozems and an increase of spatial extent of Chernozems and Kastanozems.},
  language  = {en}
}