Articles | Volume 9, issue 15
Atmos. Chem. Phys., 9, 5575–5586, 2009
Atmos. Chem. Phys., 9, 5575–5586, 2009

  06 Aug 2009

06 Aug 2009

Comparison of conventional Lagrangian stochastic footprint models against LES driven footprint estimates

T. Markkanen*,1, G. Steinfeld3, N. Kljun4, S. Raasch3, and T. Foken1 T. Markkanen et al.
  • 1Department of Micrometeorology, University of Bayreuth, 95440 Bayreuth, Germany
  • 3Institut für Meteorologie und Klimatologie, Leibniz Universität Hannover, Hannover, Germany
  • 4Department of Geography, Swansea University, Singleton Park, Swansea, Wales UK
  • *current address: Finnish Meteorological Institute, Climate and Global Change Research, P.O. Box 503, 00101 Helsinki, Finland
  • **current address: Department of Physics, Carl von Ossietzky University of Oldenburg, Oldenburg, Germany

Abstract. In this study we introduce a comparison method for footprint model results by evaluating the performance of conventional Lagrangian stochastic (LS) footprint models that use parameterised flow field characteristics with results of a Lagrangian trajectory model embedded in a large eddy simulation (LES) framework. The two conventional models follow the particles backward and forward in time while the trajectories in LES only evolve forward in time. We assess their performance in two unstably stratified boundary layers at observation levels covering the whole depth of the atmospheric boundary layer. We present a concept for footprint model comparison that can be applied for 2-D footprints and demonstrate that comparison of only cross wind integrated footprints is not sufficient for purposes facilitating two dimensional footprint information. Because the flow field description among the three models is most realistic in LES we use those results as the reference in the comparison. We found that the agreement of the two conventional models against the LES is generally better for intermediate measurement heights and for the more unstable case, whereas the two conventional flux footprint models agree best under less unstable conditions. The model comparison in 2-D was found quite sensitive to the grid resolution.

Final-revised paper