A study of two mass-consistent models: Problems and possible solutions

Two existing mass-consistent models, COMPLEX and NOABL, were tested in three regions in the UK: Devon, the Northern Pennines and Shetland. In order to solve problems arising with both models, a number of modifications were made, leading to the development of a new model, MC-3. First, the entire wind field was scaled down (or up) by a certain percentage to make the wind predictions at the predictor stations as close to the observed values as possible. Second, modifications to the non-divergent process used in COMPLEX were made. Site roughness-length and anemometer height corrections were also made. The new model worked well in terms of mean wind speed predictions, and gave a significant improvement over the predictions of COMPLEX and NOABL.