CTDMPLUS: A Dispersion Model for Sources near Complex Topography. Part II: Performance Characteristics

James O. Paumier Computer Sciences Corporation, Research Triangle Park, North Carolina

Search for other papers by James O. Paumier in
Current site
Google Scholar
PubMed
Close
,
Donna J. Burns Computer Sciences Corporation, Research Triangle Park, North Carolina

Search for other papers by Donna J. Burns in
Current site
Google Scholar
PubMed
Close
, and
Steven G. Perry Atmospheric Sciences Modeling Division, National Oceanic and Atmospheric Administration, Research Triangle Park, North Carolina

Search for other papers by Steven G. Perry in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

The Complex Terrain Dispersion Model (CTDMPLUS), described in Part I of this paper, was evaluated using the SO2 field-study data from the Lovett generating station in southeastern New York state. For perspective, CTDMPLUS estimates were also compared with those from the regulatory version of the Rough Terrain Diffusion Model (RTDM). For comparisons unpaired in space or time, the highest 25 CTDMPLUS model predictions tended to overpredict the highest 25 hourly observations by, on average, about a factor of 2. Similar results were found for 3-h and 24-h avenge predictions. Overpredictions occurred mainly for stable atmospheric conditions.

In contrast, the hourly and 3-h average model concentrations paired in time with observations underpredicted the observations by as much as a factor of 4. CTDMPLUS displayed no strong bias in estimating the 24-b average concentrations.

To understand the performance of CTDMPLUS, the meteorological conditions associated with the highest 25 observed concentrations were examined. This analysis suggests that the most significant factors affecting CTDMPLUS predictions for stable conditions are the height of the plume and its relation to the dividing streamline; in convective conditions, significant factors are the fraction of plume material penetrating the stable layer aloft, lateral plume spread, and wind direction.

Abstract

The Complex Terrain Dispersion Model (CTDMPLUS), described in Part I of this paper, was evaluated using the SO2 field-study data from the Lovett generating station in southeastern New York state. For perspective, CTDMPLUS estimates were also compared with those from the regulatory version of the Rough Terrain Diffusion Model (RTDM). For comparisons unpaired in space or time, the highest 25 CTDMPLUS model predictions tended to overpredict the highest 25 hourly observations by, on average, about a factor of 2. Similar results were found for 3-h and 24-h avenge predictions. Overpredictions occurred mainly for stable atmospheric conditions.

In contrast, the hourly and 3-h average model concentrations paired in time with observations underpredicted the observations by as much as a factor of 4. CTDMPLUS displayed no strong bias in estimating the 24-b average concentrations.

To understand the performance of CTDMPLUS, the meteorological conditions associated with the highest 25 observed concentrations were examined. This analysis suggests that the most significant factors affecting CTDMPLUS predictions for stable conditions are the height of the plume and its relation to the dividing streamline; in convective conditions, significant factors are the fraction of plume material penetrating the stable layer aloft, lateral plume spread, and wind direction.

Save