| dc.description.abstract | Soil moisture plays a crucial role in land-atmosphere interactions through its ability to divide the incoming
radiation into latent heating and sensible heating. Accurate modelling of soil moisture could significantly
improve numerical weather prediction, but observational data is difficult to obtain. In this research, the Tiled
ECMWF Scheme for Surface Exchanges over Land (TESSEL) is coupled to Whiffle’s large eddy simulation model
called GRASP and the behaviour of this system is explored. It is concluded that TESSEL is highly sensitive to
initial conditions of soil moisture and soil parameters. In 6 months of model runs, rain-soil moisture coupling
improves the model results through increased latent heating. Differentiation between soil types in Cabauw and
Gilze-Rijen also increases the model skills, concluded from decreased biases and increased correlations with
observations of surface fluxes, temperature, humidity and wind speed. This effect is strongest in summer and
also influences wind speed and cloud formation. In a radiation fog case study, initialization with atmospheric
observations produced a fog layer close to observations in depth, with the dissipation delayed by an hour. For
this model set-up, it is recommended to permanently add differentiation between soil types, so that the initial
soil moisture matches the soil parameters. Some improvements made to TESSEL should be added such as
the conductivity formulation, the bare soil evaporation and the addition of surface run-off. In future research
the behaviour of the system during heavy rainfall could be studied, as well as the influence of open boundary
conditions and the use of other data sources for initialization. | |
