Relative Bias For ¢ither wind system, relative bias may be defined by two bias para- meters, bias direction and bias ratio. Bias direction describes the orien- tation of the collection geometry and is the angle measured clockwise from the reference direction (bow or true north) to the minimum-maximum axis of the collection geometry or pattern. Bias ratio is a measure of the magnitude of relative bias and is defined as the ratio of the maximum to the minimum value of the collection curve which represent the variation of collection around the platform. The bias direction of a s.w. system is merely the wind direction and this fact serves as a criterion in the orientation of s.w. collection pattern or curve. For this system the bias ratio increases with increasing wind speed and decreases with increasing particle size and density. in the mw. system, wind velocities and the relative amount of fallout associ- ated with each wind velocity must also be taken into account in the variation of bias direction and bias ratio. xt is recognized that particle shape is another variable to be considered not only because of the aerodynamic effects but it also may be an indication of different particle types with intrinsic differences in the concentrations of radionuclides in the particles.> However to maintain simplicity in diseussion of certain basic bias relationships, this variable is not included in this study. The obje:tiove of relative bias analysis is the detezmination of the eollerzticn curve* from which the describing bias parameters and a significant mean platform value may then be derived. As will be shown, the number of sample values per plstform were insufficient to adequately describe the collection curve; hence, interpolation has been used extensively. To aid in thie inter;lstion, the results of an air flow study have been used in conjunction with the actual collection data tc establish the important characteristics of relative bias. Air Fiow Studies To investigate the air fiow characteristics above the standard platform in @ given wind, wind tunne!. model studies, as well as smoke and wool tuft *It is to be noted that the intrinsic efficiencies of the collectors are not known and therefore sample values may not represent absalute amounts deposited; nevertheless the values do indicate relative bias since collectors of identical efficiencies were employed.