*a + ed el . .k 2 4* 7 pel ewtal ice dtlalid tete Ta te ce ee ses oe wet 160 ventory, including point as well as area sources; (2) the network of meteorological stations; (3) the network of pollutant monitoring stations; and (4) the mathematical algorithm which describes the processes which transform the concentrations at the source to those observed at the receptor. Also, a substantial amount of attention is devoted to verification procedures. Two general categories of mathematical models are discussed. Thefirst is the source-oriented model characterized by the work of Turner. In this model, the concentration at a particular point is determined by the superposition of the contributions from each of the sources upwind. The second is the receptor-oriented mode! of which there are three types: the first is that described by Clarke, in which the pollutant concentrations at a particular monitoring station are analyzed in a manner similar to that of the source-oriented model. Second, there is the regression model, in which regression equations are developed relating observed concentrations at a receptor with meteorological variables; and third, the tabulation prediction scheme in which combinations of meteorological variables are arranged in an ordered sequence. For each combination, the probability distribution of pollutant concentration is given along with other statistical pa- rameters, such as the interquartile range, the mean, and the standard deviation of the distributions. With the tabulation prediction scheme, one may look up the meteorological conditions just as one looks up a name in a telephone book or a word in a dictionary and read off the probability distribution of concentrations. A summary of verification techniques is given with examples of how other workers in the field have veri- fied their models. These include scatter diagrams, cor- relation coefficients, and isopleth comparisons. The weaknesses and strengths of validation techniques are discussed. THE USE OF PYRHELIOMETERS FOR CONTINUOUS MEASUREMENTS OF AN EFFECTIVE AIR POLLUTION MIXING DEPTH Harry Aloses and D. N. Eggenberger* Information on the magnitude of the mixing depth is important for forecasting levels of pollutant concentrations over a given area. To date, mixing depth information has been obtained by means of sensors mounted in an aircraft or carried aloft by balloons. At best, the information provided has been sporadic. This paper describes techniques using combinations of either pyrheliometers or pvranometers to provide continuous near the ground to several thousand feet. The covariance between the signals recorded by the two instruments would change as the common volume passed from air with a high concentration of aerosol to clean air. By noting the height at which the changes occur, it should be possible to determine both the height of the aerosol mixing layer and also the presence of aerosol layers above the primary ground base lavers. recordings of the mixing depth. Various techniques are de- scribed for obtaining these measurements. The most attractive of these consists of two pairs of pyrheliometers on equatorial mounts; one pair is located on a tall building and another pair near the ground. One instrument of each pair has a 40004500 A filter and the other. a 5500-6000 A filter. By considering ratios of the solar radiation recorded by combinations of these pytheliometers, it is posstble to determine the depth of the aerosol mixing layer under the assumption that the aerosol concentration is approximately uniformly mixed within the mixing layer and drops to a very small value aboveit. Variations from this aerosol distribution can be handled when three or four pairs of instruments are used at various heights with the greatest height approaching 1000 ft. ; The correlation technique involving the crossing of two pyranometer fields of view is also discussed. In this method, one pyranometer has a conical field of view of about % degree directed upward, The second pyranometer, also with a narrowfield of view, is located about 1000 to 2000 ft away. The fields of view of the two pyranometers are made to intersect. The second pyranometer is designed to scan by changing its elevation angle so that the height of a common volume of the two intersecting fields of view varies from a level * Electronics Division. INTRODUCTION Whenone approaches an industrialized urban area elther byaircraft or by automobile, a pall of pollution is readily discernible. The emission of pollutants into the urban atmosphere coupled with meteorological processes such as diffusion, transport, or convection controls the height of this pollution or atmospheric mixing layer. One may consider the mixing layer in different ways: (a) the vertical distribution of pollutants, (b) the verti- cal distribution of temperature, or (c) the vertical wind profile. During steady state conditions, one expects agreement among these three types of measurements. During dawn or dusk—transitional pertods—appreciable differences may be observed because of different adjustment rates or system time constants. If the mixing depth is small, levels of pollutant concentrations are high since the pollutants are mixed with air confined to a narrow layer; with a large mixing