In this study, a North American Regional Reanalysis (NARR)-based climate dataset of wind and Pasquill stability class is developed for the North American Great Lakes region. Motivated by the need to update the limited wind and Pasquill stability class climatology previously incorporated into a livestock odour nuisance-mitigation application used in Michigan, this study is motivated more generally by the lack of a long-term climate dataset suitable for studies of atmospheric dispersion in the Great Lakes region. As a preliminary step, NARR 10-m wind speed, 10-m wind direction, and 2-m temperature are evaluated at eight weather stations in Michigan and northern Indiana, and the wind speed variable is subsequently bias-corrected. To derive a Pasquill stability class dataset from the NARR variables, an existing empirical relationship between Pasquill stability class and Richardson number is applied to a previously developed NARR-derived Richardson number dataset covering the period 1979–2008. Pasquill stability class frequency in NARR is subsequently evaluated using observations at two flux towers in Illinois. Although a tendency is noted for the NARR-derived Pasquill stability class dataset to underestimate the frequency of stable and unstable conditions, and overestimate the frequency of neutral conditions, the analysis shows that the 10-m wind speed bias-correction procedure yields a frequency distribution that is in better agreement with the observations than the non-bias-corrected dataset. Finally, an analysis of the spatial and temporal variability of livestock odour dispersion potential in Michigan is provided as an example application of the NARR-based wind and Pasquill stability class climate dataset.
Read full abstract