Dependence of deposition velocity on the frequency of meteorological observations for the Chesapeake Bay

Abstract

Abstract Williams' (1982) model was used with meteorological data from a NOAA buoy station in the northern Chesapeake Bay to investigate the behavior of deposition velocity (V d ) vs. particle size curves determined for periods of 1 day and 1 week using various averaging times and observational frequencies. The model predicts a nearly 200-fold difference in the V d of a 0.5 μm particle for wind speeds of 2 and 20 m·s −1 , which suggests that the wind speed frequency distribution must be considered when applying V d s to concentration data for discrete sampling periods. The results for 0.5 μm particles determined for 5 July, 1990, a day with average winds of 5.5 m·s −1 and 1 hourly observation of 16 m·s −1 , differ by 2-fold when the model was run with 24 one-hour average observations of wind speed and air/water temperature, and, again, with 1 observation representing the average of the 24 hourly observations. When the model was run for the 1-week period, 4–10 July, with 7, 14, 28, 42, and 168 hourly observations of wind speed and air/water temperatures the results for all simulations differed by ≤ 14%. However, this time period contained only one occurrence of wind speed > 15 m·s −1 . Clearly the frequency and intensity of gusts must be considered when estimating dry deposition fluxes for atmospheric pollutants.