Methodology I – Overview
Facts about Mesoscale and Microscale Wind Modeling as Tools for Wind Resource Assessment and Mapping
Mesoscale and microscale wind modeling are two computer based techniques for simulating, with reasonable accuracy, complex wind flows in areas where surface measurements are sparse or non-existent. Both tools provide an assessment of the wind resource and subsequent calculation of the annual energy output of wind turbines.
Model size, resolution and topography
Both modeling techniques are used individually or coupled, depending on the size of the model area, the targeted model resolution and the complexity of the underlying topography.
Meteorological centers (such as CMC, NCEP or ECMWF) collect and analyze, every six hours, the data from surface stations, radiosondes, ships, airplanes, radar installations, and satellites. They provide long-term (several decades) quality-controlled and analyzed 3D weather data covering the world. However, the horizontal resolution of the data is too low (on the order of hundreds of kilometers) to be directly used for wind farm siting.
Mesoscale simulations are best suited to estimate the wind situation on a regional scale (reproducing for example the Foehn wind or the Mistral in the computer model) whereas microscale simulations are better suited to estimate local wind systems and turbulences.
scale and micro-
The statistical-dynamical downscaling method  is applied when coupling mesoscale and microscale modeling. The downscaling method involves the computation of a mesoscale model based on large scale, long-term meteorological data and their statistical properties. Statistical post-processing of the mesoscale modeling results produce mesoscale wind statistics. Microscale modeling is then used to refine the mesoscale results, adjusting the regional wind resource with high resolution data of topography and roughness.
Microscale modeling is based on two different computational approaches using linear and non-linear equations. The use of non-linear equations to describe the state of the atmosphere results in a higher degree of accuracy when compared to using linear equations, in particular for areas with steep terrain slopes (more than 16-17°) and complex topography. Higher accuracy involves increased computation time and usually smaller model areas. Models using linear equations are favored for model areas with simple to moderate terrain slopes.
Wind is influenced by the topography and land cover. These geophysical features are required by the mesoscale and microscale models and have to be described for the computational domain. Topographic and land cover data are available in databases from different sources.