CTSCREEN Scaling Factors for the Tropics

In 1991, the U.S. Environmental Protection Agency made the CTSCREEN model available to calculate a worst case 1-hour concentration in complex terrain. In order to determine a concentration for longer averaging times, scaling factors were developed. This was done by determining the ratios between the 1-hour CTSCREEN concentration and the CTDMPLUS modeled concentrations which were based on actual meteorological data. At the time, only a limited number of CTDMPLUS type meteorological data sets existed for the purpose of developing scaling factors. The data sets were representative of continental United States locations and included Westvaco (New Jersey), Lovett (New York) and Widows Creek (Alabama). None of these data sets were representative of tropical locations such as Hawaii and Puerto Rico. Unlike the continental United States, the tropics have greater wind direction persistence due to trade winds.

This study consisted of two objectives. The first objective was a sensitivity analysis to substantiate the existing CTSCREEN scaling factors using continental United States meteorological data sets from Westvaco, New Jersey, Martins Creek, New Jersey, and East Helena, Montana, and tropical meteorological data sets from Guayama, Puerto Rico and Kahe Point on the island of Oahu, Hawaii. Original and new topographical features and stack configurations were used in this objective. The second objective was a derivation of tropical scaling factors using the Hawaii and Puerto Rico meteorological data sets, and local terrain and various stack configurations typical of Hawaii.

Study results confirm the 1991 CTSCREEN 3-hour, 24-hour and annual average scaling factors as being reasonable. However, the tropical scaling factors were found to be greater than the continental United States scaling factors. Consequently, a set of scaling factors for the tropics has been identified for use in this paper. Because the CTSCREEN model did not contain 8-hour and quarterly scaling factors, this paper also identifies continental United States and tropical scaling factors for these two averaging times.