Improvement of an aerodynamic roughness model with meteorological measurements and TM image

Aerodynamic roughness length (z0) is one of the main parameters for describing aerodynamic characteristics and turbulence exchange of underlying surfaces. It also plays a very important role in land surface fluxes simulation and atmospheric boundary layer studies. In fluxes and atmospheric numerical simulation models, z0 is usually considered as a simple parameter related to rough elements, while actually z0 is the result of aerodynamic and thermal factors and rough elements. Some researchers on z0 established a model to calculate effective roughness at TM pixel scale. But in the model, the length of the fetch is simply considered as 100 times of the highest measuring height, which results in much uncertainty of the model. With the development of eddy covariance observations and the footprint theory, fetch could be determined with more accurate footprint model. This paper aims to make an improvement for the model basing on footprint model, and clarify the relationship between z0 and air temperature, wind speed and geometric roughness. Firstly meteorological data in two different levels was obtained to calculate z0 and the calculated z0 was considered as the real value, and then the fetch length was calculated by FSAM footprint model. Secondly LAI was retrieved from TM image to calculate geometric roughness. Thirdly drag coefficient was taken into the roughness calculation model, and then a roughness calculation model was established with footprint model and LAI, vegetation height, atmospheric stratifications and drag coefficient. The result indicated that z0 is the function of thermal and aerodynamic factors, and the reason for roughness dynamic was interpreted thoroughly. The quantitative relationship between rough elements and aerodynamic and thermal factors is clarified, and a more accurate model for retrieving z0 at TM pixel scale could be established with the fetch determined by footprint.