United States Environmental Protection Agency Environmental Sciences Research Laboratory ~ " Research Triangle Park NC 27711 Research and Development EPA-600/S3-82-042 Oct. 1982 Project Summary Determination of Dry Deposition Rates for Ozone Elmer Robinson, Brian Lamb, and M. P. Chockalingam Measurements were taken of the rates of loss of Os from the atmosphere onto the earth's surface under non- precipitation conditions. This is com- monly called "dry deposition" and the critical parameter is usually the "deposition velocity" Vd. Other in- vestigators have determined that the general magnitude of Va for O3 is about 0.6 cm/s. The major objective of the present research was to measure ozone Vd and other meteorological parameters over a wide range of ambient conditions and to relate the values of Vd to commonly measured or estimated meteorological parameters. In this way a calculation procedure for ozone Vd cou Id be developed for use i n meteorological pollutant transport models. The micrometeorological profile method for estimating Vd was used in this research program. Measurements were made over a soybean field in Illinois, a grain and grass field in Pullman, WA, and a hardwood forest area in south central Pennsylvania. The measurements over the soybean and grain fields show that Vd has a strong diurnal cycle and is correlated with a lateral gustiness parameter, a u, where og is the standard deviation of wind direction and u is the average wind speed. The model is of the form Vd = A(OQ u)b where the value of A is between 0.2 and 0.4 and b is between 0.7 and 1.3. The measurements of Vd over the Pennsylvania forest canopy did not follow this model, but the reasons are not clearly understood. However, there were some notable meteorological features of the experi- mental period. In particular, there was little diurnal variation in turbulence over the forest; conditions were generally turbulent over the forest canopy during the day, as might be expected, but also, unexpectedly, at night. Values of Vd over the forest area were generally larger by a factor of two or more compared to the Vd measurements over the crop surfaces. This Project Summary was developed by EPA's Environmental Sciences Research Laboratory. Research Triangle Park, NC. to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). Introduction Air pollutants are scavenged from the atmosphere largely by adsorption or reaction at the earth's surface. The effects of scavenging mechanisms usually have not been considered important factors in plume diffusion modeling because the models have been applied to relatively short distances; thus, the time periods being considered would have been too short for deposition to have had a significant impact on the system. However, diffusion and trans- port are now being modeled over much greater distances and covering longer time periods. Under such conditions it is no longer prudent to ignore the sca- venging phase of the transport model. The concept of deposition velocity (Vd) was first described in a 1953 paper by A. C. Chamberlain and R. C. Chadwick from the Atomic Energy Research ------- Establishment, Harwell, Great Britain, as a means of explaining the loss of material from the atmosphere to the underlying surface when the material was too small to be covered adequately by Stokes' law. By definition, Vd is the coefficient that relates air concentra- tion, C, to the deposition rate, D: D = VdC where the deposition rate, D, is essen- tially the downward flux of the pollutant (fjg cm "V1); and the concentration of pollutant, C, can be expressed in //g/cm3. Hence, the dimensions of Vd must have the units cm/s in order for the expression to balance. Since these are the dimensions of velocity, it was natural to label the coefficient deposition velocity. Further study has shown that the surface loss is equal to the downward flux to the surface, f, which can be expressed in terms of well known micrometeorological variables and the vertical concentration profile. The primary objective of the experi- mental program was to measure the dry deposition velocity of Oa. A second objective was to relate the deposition velocity for Oa to a wind or a gustiness parameter, such as ad, for inclusion in Gaussian-type plume models. The profile method was selected as the Vd estimation technique. The field measurements were con- ducted in several major phases at three locations. In July and September, 1977 measurements were made at Robinson, IL. This was followed by a program in Pullman, WA from April to July, 1978 and near Lancaster, PA in July-August, 1979. The underlying surfaces at these three locations were different. In Illinois, the measurements involved a soybean field at two stages in its growth cycle; in Washington, the surface was grain and grass. The measurements in Pennsylvania were made over the top of a deciduous forest with a 20-m high canopy. The measurements produced an extensive set of dry deposition and meteorological data that can be applied to modeling programs. Procedure The field system consisted of four major components: (1) an instrumenta- tion tower of either a scaffold type or a taller antenna type; (2) the meteorolog- ical sensor system; (3) the air sampling system; and (4) a field facility to house the instrumentation. The Robinson IL field study began July 15, 1977 with the installation of meteorological and air sampling instru- ments adjacent to the Robinson airport, and continued to July 31, 1977. A similar experiment was conducted at the same site from September 5 to September 15, 1977. The scaffold instrumentation tower was erected at the eastern edge of a soybean field. Because flat agricultural lands extend- ed several km upwind to the west, winds blowing from the directions of south through west to north were generally applicable to profile experiments; this was especially true for winds from the sector enclosed between 225° and 360°. The most frequent winds were from westerly directions satisfying the experimental requirements that the wind and pollutant flow be stabilized over an "infinite uniform" surface before reaching the tower location and profile sensors. Commercial steel scaffold compo- nents were used to construct the scaffold instrumentation tower. Mete- orological sensors weref ixed at levels of approximately 25, 75, 200, and 500 cm above the crop surface. The meteoro- logical sensors and air sampling intakes were located about 2 m from the structure in the direction of the prevail- ing wind. The observation levels and the types of observations on the tower are presented in Table 1. The second phase of the study was conducted between April 1 and August 7, 1978 on a site approximately 4.8 km northwest of Pullman, Wa. At this site it was necessary to erect the scaffold instrumentation tower on the gentle slope of a hill which forms a shoulder of a shallow valley in the midst of rolling hills. The prevailing wind was from the southwest along the axis of the valley; this satisfied in a reasonable manner the requirements that the windflow be stabilized over an "infinite uniform" surface before reaching the instru- mentation. Although low hills with gentle slopes rising about 30 m above the valley axis were located about 0.8 km southwest of the tower, the general terrain conformed more or less to a uniform formation which ensured a stable fetch of wind for a distance of approximately 8 km to 16 km. The crop surface was a mixed growth of grain, grass, and weeds with an average crop height of approximately 30 cm at the beginning of the April 1978 experi- ments. The same scaffold tower instru- mentation profile given in Table 1 was used for these experiments. The Pennsylvania forest site was located on Tower Mountain between Lancaster and York. The tower was set up at the edge of a predominantly oak forest with an estimated canopy height of approximately 20 m. To the west, in the direction of the prevailing wind, the forest surface was relatively uniform for about 2 km. The ground was cleared and leveled at the edge of the forest where the tower was located. A 34-m antenna tower was assembled on the ground with the instruments and sampling lines in place. It was then tilted into position using a gin pole and winch with the supporting guy cables used to maintain a stable attitude during the lift. The profile extended approx- imately 12m above the top of the canopy. The observation levels and types of observations made at the forest site are presented in Figure 1. At all three experimental sites air samples for O3 analysis were drawn from six levels above the canopy surface Table 1. Scaffold Tower Instrumentation Profile Observation Level Profile Height* Type of Observation 1 2 Crop Surface fapprox.J 0.25m 0.75 m 2.00m 5.00 m Air Sample Inlet Air Sample Inlet, Wind Speed, (WSJ, Temperature, (TEMP.), Dew Point (DP) Air Sample Inlet, Wind Speed, Wind Direction, fWDJ, Direction Sigma, la*). Temperature Air Sample Inlet, Wind Speed, Temperature Air Sample Inlet, Wind Speed, Wind Direction, Direction Sigma, Temperature, Dew Point *Profile heights are given in relation to the top of the underlying vegetation. ------- 34.0 32.0- § I I! o •Q 25.8- 22.6- 20.9- 20.0- 7 1 7 7 WS,WD,Oa. Oe -A-WS,Temp,O3 WS,WD,Temp,03.0e "7"'\WSJem~p~0 Figure 1. Antenna tower instrumentation for forest profile studies, Pennsylvania, 1979. using 0.5 in O.D. (0.25 in, I.D.) Teflon tubing through a solenoid valve assem- bly into a common manifold leading from the tower to the mobile laboratory. A pump inside the mobile laboratory moved the sample air stream through the common manifold and the pollutant analyzer obtained its sample from this manifold. By electrically sequencing and opening one valve at a time, it was possible to draw an air sample into the common manifold successively from each of the six sampling ports. Each valve was kept open for 5 minutes. Line loss of Oa through the manifold was determined and suitable corrections were applied to the measured concen- trations. Results and Discussion Crop Sites The experimental data from the Washington and Illinois sites for ozone Vd values apply to a height of about 1 m above the crop surface and show a general correspondence between the three experimental periods. At night, values of Vd were often close to zero except when the nighttime winds maintained some turbulence and pre- vented the formation of a strong radiation inversion. The typical daytime Vd value of approximately 0.6 cm/s is similar to recent values published by investigators in England and the United States. The nighttime Vd range of 0.1 cm/s to 0.3 cm/s, under situations when the surface-based nocturnal inversion did not occur, agrees well with the range reported from measure- ments in England. The trend of variation of the ground level Os concentration generally exhibits a diurnal cycle with clearly defined daytime maxima and nighttime minima at both crop sites. This is a normal situation and has been reported by other investigators. It is attributed to the transport of Os within and below the nocturnal inversion and destruction by deposition onto the underlying surface. The morning increase of Os concen- tration is attributed primarily to downward mixing of air from aloft when the radiation inversion is dissipated. In some situations, photochemical Oa formation could play a role in morning concentration patterns. At the Illinois and Washington sites, the vegetation characteristics were generally uniform for several hundred meters in the direction of the prevailing wind. There is no evidence that the terrain features adversely affected the profile measurements. The 03 dry deposition velocity and Os flux were calculated using only profiles recorded during the time periods when the wind fetch was from a suitable sector. The recorded profiles of wind speed, temperature and Os concentrations were considered for an averaging period of 30 min. A graphical procedure was followed in obtaining the average profile for Os concentration during the 3 min averaging period. A similarity between the transfer of heat and the transfer of pollutant was assumed in relation to ozone Vd and ozone flux estimates. The Os concentrations in these cropland experiments were indicative of "natural" or background conditions rather than urban areas photochemis- try. This was perhaps a significant difference compared to the forest experiment. The field data show a general corre- spondence between the three experi- mental periods and the two sites. Daytime values of Vd averaged 0.6 cm/s with a standard deviation of 0.17 cm/s. At night Vd values were in the general range of less than 0.1 cm/s to 0.3 cm/s, with some higher values when the nighttime winds were especially strong. The daytime average Os flux deter- mined from the cropland field study data is around 6 x 1011 molecules cm"2 s"1, and the nighttime flux is about 1.5 x 10" molecules cm"2 s"1. The Os flux reported * U S. GOVERNMENT PRINTING OFFICE 1062-559-017/0834 ------- by other investigators is in the general range of 1 x 10" molecules crrf2s~1 to 6 x 10" molecules cm~2 s"1. From the basic considerations of the micrometeorology of the boundary layer, we would expect Vd to be generally related to the product G of lateral wind gustiness, erg , and the mean wind, u. From data at the two sites, acceptable statistical relationships between ozone Vd and G were obtained as follows: Robinson, IL: July, 1977: Vd = 0.0339(G)084 September, 1977: Vd =0.0241 (G)078 Pullman, WA: April/May, 1978: Vd = 0.0406(G)076 where G = OQ u (deg m s 1). Although there are some differences between these three power law regres- sions that may be due to differences in the surfaces, they are still generally similar and have been combined into a single power law regression: Vd =0.0322(G)079 This expression could be used to derive values of Vd from estimates of og and u for a variety of cropland surfaces and sites. Forest Site The forest experiment produced data that were not as easily interpreted as the Illinois and Washington crop site data. The top of the forest canopy apparently did not develop a strong diurnal pattern of nighttime stability and daytime turbulence, and there was a great deal of scatter in the values of all the calculated parameters. On the basis of median values of the various param- eters, the boundary layer over the canopy was determined to be charac- teristically unstable and turbulent. Median values of Vd were about 1 cm/s with no significant diurnal cycle Although the gustiness parameter, erg U, had an apparent diurnal cycle with maximum median values in the late afternoon, nighttime median values were also relatively large The correlation between Vd and erg u was much lower than for the crop experiments. It is believed that the Vd estimates showed relatively large fluctuations because the Os profiles were not in equilibrium with the canopy surface. This probably resulted from local formation and scavenging processes due to the influ- ences of local and regional photo- chemical air pollutants. It appears that the model developed from the low crop experiments underestimates values of Vd for the forest site by 50 percent or more. Recommendations The application of these results to transport model calculations was a principal objective of this research and it is believed that the results described above provide reasonable process toward this goal. For large-scale trans- port studies, and in most other model studies, it will be necessary to estimate values of erg and u from the prevailing meteorological conditions. Values of u can be obtained directly from the observed or forecast wind field. Esti- mates of erg can be obtained from an estimate of surface layer stability and the relationships between stability and erg. The stability classes can be deter- mined from general weather conditions using generally accepted techniques. Elmer Robinson, Brian Lamb, and M. P. Chockalingam are with Washington State University, Pullman, WA 99164. W. A. Lonneman is the EPA Project Officer (see below). The complete report, entitled "Determination of Dry Deposition Rates for Ozone," (Order No. PB 82-258 989; Cost: $10.50, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Environmental Sciences Research Laboratory U. S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Postage and Fees Paid Environmental Protection Agency EPA 335 Official Business Penalty for Private Use $300 0000329 46Ehcl ------- |