Measured and calculated heat flux above a freezing and melting ground

Three turbulence instruments were used in a field experiment, WINTEX that was performed in March 1997. The site was located 10 km north of Uppsala, Sweden. The instruments used were a Scintec Scillentometer, a hot wire instrument and a Solent sonic anemometer. The two first instrument were located at 2m above ground and the sonic was located at 10m above ground. The ground was covered with grass during the campaign and there was no snow on the ground. Profile measurements indicate that there was a strong convective layer close to the surface for some of the days.

During the measurements the thermal state of the ground were changing. For the first period of the measurements the soil at the site was unfrozen and the temperature fluctuations in the soil were following the pattern of the temperature in the lower air layers. During the middle period the soil became frozen, the temperature of the soil in the upper layers were around 0 °C. while the temperature of the lower air layers had a diurnal amplitude of the order of 10 °C. At the end of the period the upper layers of the soil were frozen and a diurnal variation could be observed. During the daytime there were melting processes in the upper layers, taking care of a lot of the incoming heat from solar radiation. The smallest heat fluxes occurs during the first period when the surface was heated and a lot of the energy was transported downwards to lower depth.

Numerical calculations have been performed by two different kinds of models for the studied period. A soil model has been used for studies of the processes in the soil. For the atmospheric processes a meso scale model has been used.

The conclusion from this study is that the state of the soil is important for the structure of the heat flux. In this case study the under saturated ground were de-coupled in two layers, by a dry section (no water between the soil particles), when the lower part of the soil was frozen. The observed heat flux during this situation were higher than when the soil was not frozeen and the incomming radiation was higher. The results from the numerical calculations verify the results from the measurements.

The data were also used for determining in what extent the soil status has to be implemented in a meso scale model. The result from this study is that the cases when the soil is going through phase changes have to be treated carefully. In the other cases a not so sophisticated parameterisation of the ground is sufficient in order to generate sensible heat flux of the right magnitude. The general conclusion is that for the situation described above a complete soil model has to be used in order to generate the surface temperature which than can be used as lower boundary condition in a meso scale model.