Articles | Volume 4, issue 2
Earth Syst. Dynam., 4, 187–198, 2013
https://doi.org/10.5194/esd-4-187-2013
Earth Syst. Dynam., 4, 187–198, 2013
https://doi.org/10.5194/esd-4-187-2013

Research article 11 Jul 2013

Research article | 11 Jul 2013

A stochastic model for the polygonal tundra based on Poisson–Voronoi diagrams

F. Cresto Aleina1,2, V. Brovkin2, S. Muster3, J. Boike3, L. Kutzbach4, T. Sachs5, and S. Zuyev6 F. Cresto Aleina et al.
  • 1International Max Planck Research School for Earth System Modelling, Hamburg, Germany
  • 2Max Planck Institute for Meteorology, Hamburg, Germany
  • 3Alfred Wegener Institute for Polar and Marine Research, Research Unit Potsdam, Potsdam, Germany
  • 4Institute of Soil Science, Klima-Kampus, University of Hamburg, Hamburg, Germany
  • 5Deutsches GeoForschungsZentrum, Helmholtz-Zentrum, Potsdam, Germany
  • 6Department of Mathematical Sciences, Chalmers University of Technology, Gothenburg, Sweden

Abstract. Subgrid processes occur in various ecosystems and landscapes but, because of their small scale, they are not represented or poorly parameterized in climate models. These local heterogeneities are often important or even fundamental for energy and carbon balances. This is especially true for northern peatlands and in particular for the polygonal tundra, where methane emissions are strongly influenced by spatial soil heterogeneities. We present a stochastic model for the surface topography of polygonal tundra using Poisson–Voronoi diagrams and we compare the results with available recent field studies. We analyze seasonal dynamics of water table variations and the landscape response under different scenarios of precipitation income. We upscale methane fluxes by using a simple idealized model for methane emission. Hydraulic interconnectivities and large-scale drainage may also be investigated through percolation properties and thresholds in the Voronoi graph. The model captures the main statistical characteristics of the landscape topography, such as polygon area and surface properties as well as the water balance. This approach enables us to statistically relate large-scale properties of the system to the main small-scale processes within the single polygons.

Download
Altmetrics
Final-revised paper
Preprint