Determining the best CALMET option settings to produce windfields in the surface layer

The surface windfields produced by the CALMET meteorological model have been evaluated for accuracy by comparison to an extensive data set of observations compiled for the summer of 1999 for a domain in northern New England (Riley, 2002, AWMA Publication). The observational data set consisted of surface measurements from the ASOS sites in the National Weather Service network as well as data from a few sites in a local MESONET network. The evaluation computed directional and speed error for the CALMET windfield predictions paired in space and time to the observation data set predictions. These error estimates were categorized for a range of meteorological conditions and site characteristics for the observational sites. In the evaluation the CALMET model was run with many different combinations for values of the parameter settings that most affect windfield generation in complex terrain in an effort to determine the best settings for use on the domain. The evaluation was performed for CALMET runs at a horizontal grid resolution of 0.2 kilometers. The evaluation also examined the acceptability of usage of these windfields for long-term air quality standards through temporally independent assessments of model accuracy. Most significant final conclusions are that the model underestimates speeds and that the directional error of the model is much less at higher windspeeds. Inversely it may be concluded that the National Weather Surface observations represent wind speeds higher than those generally occurring in windfields over complex terrain because the measurements are taken over airport plains.

In order to ascertain performance of the CALMET model under different meteorological conditions which affect air pollutant dispersion and mixing, hours during this summer period were each characterized by one of twelve different meteorological categories. Meteorological categorization was assigned for each hour at each site independently. The meteorological categories assigned to each site-hr differ by wind speed, thermal stability, and dewpoint temperature criteria. Wind speed and dewpoint were determined from the ASOS data itself, and thermal stability was determined with the surface temperature and the 850 mb. temperature from Albany, NY raw insonde data. The thermal stability criteria involved making comparison to a critical lapse rate of .005 deg C / meter as an inversion sufficient to cap vertical mixing (Heffter, 1983). The categories are defined in detail in Table 1. Note that humidity is a defining variable only for the less stable categories, i.e. daytime conditions, in an attempt to define air mass origin. For stable, nighttime conditions it was assumed that the much greater relative humidity was induced by radiational cooling, and not relevant to air mass definition. In general, the potential for mixing in the boundary layer increases with the category number.

Table 1. Criteria for Meteorological Categories.

Below some overall results are given from this study after applying the meteorological categorization above.

Here is an example plot of 24 comparative runs which were conducted varying parameters sensitive to terrain effects then ordering the runs from one to twenty four based on an overall measure of model performance for each run.

Last Updated: 1/22/03