Modeling Cold Season Heat Fluxes
over an Arable Field in Central Sweden
Manfred Sthli1, David Gustafsson2, and Per-Erik Jansson2
A climate change will significantly alter the surface conditions at high latitudes, but the predictions
include many uncertainties. The energy exchange between land surface and atmosphere is not well
understood, especially not during winter conditions. Our specific purpose is to quantify the heat
exchange at the surface of an arable field in mid-Swedish conditions, during both snow and snow-
free conditions, and to link the detailed process description of the plot scale with the spatial distribu-
tion observed over a field of some hectares. Our study is a part of a currently running EU-project
(WINTEX) studying winter land surfaceatmosphere interactions of a boreal landscape at different
scales.
Extensive snow and soil physical, as well as meteorological, measurements have been set up at an
arable field station 5 km north of Uppsala (central Sweden). Measurements at selected points (pro-
files) are made of soil temperature, soil water content, and groundwater level. In addition a transect
study of snow depth is made and turbulent heat fluxes are measured with eddy-correlation technique.
The measurements have been running during winter season 1997/98. A one-dimensional numerical
SVAT (soil-vegetation-atmosphere transfer) model has been parameterized for the site and will now
be used for interpretation of the variation of the heat exchange over the field. Only vertical heat
exchange is simulated, neglecting the advective fluxes between snow-covered and snow-free areas.
Comparing these model results with eddy-correlation measurements may indicate if the strictly
vertical heat flux description is appropriate for the scale of the studied field or not.
There were mostly snow-free conditions during the winter, and only one four-week period in
JanuaryFebruary with continuously snow-covered ground. Preliminary attempts to close the energy
balance of the surface, using measured net radiation and estimated soil heat fluxes, indicate that the
eddy-correlation measurements underestimate the turbulent heat fluxes. Possible errors are the parti-
tioning between latent and sensible heat fluxes and the heat flux to/from the snow/frozen soil sur-
face. The soil heat flow is simulated using the SVAT model, which includes freezing/melting of soil
water. Simulation results will also be compared to other eddy-correlation measurement from the
same site. Much effort so far has been spent on the quality control of eddy-correlation measure-
ments, and one conclusion was that it is difficult to do these measurements during wintertime.
1 ETH Zrich Institute for Terrestrial Ecology, Grabenstrasse 3, 8952 Schlieren,
Switzerland
2 SLU, Department of Soil Sciences, Box 7014, 75007 Uppsala, Sweden
105
Previous Page
