tropical, polar, etc. Wind and weather conditions are allowed for. A particle size distribution hypothesized by the RAND Corporation is used. Aerodynamic particle fall rates are used. The method is calibrated to the AFSWP-507 reported survey of CASTLE Bravo. Be- eause of the finite number of éHecs in the model, contours are often lumpy and may involve artifact "hot spots". A problem solu- tion requires several hours by trained personnel. Air Weather Service Method The method is essentially s radex plot resulting in rough sectors within which fall-out may be expected, together with estimated times of arrival along vectors from ground zero for particular altitudes from which it is assumed particles start their fall. The method does not generate contours, and thus is not directly comparable to methods that do. The plot can be drawn in about 15 minutes by trained personnel. L Los Alamos Method The method was devised for use during CASTLE in forecasting the gross results to be expected from surface bursts of about 10 MI fired.in the northernMarshalls. by direct yield dependence. The plots are modified only The effect of yield on initial cloud geometry is not included, nor is the effect of the different tropopause height of the middle latitudes from that in the’ Marshalls. The method gives very little detail anywhere and none at all near,the origin. It is not intended for use in detailed post-shot analysis, but rather for radsafe planning. A problem solution probably requires about an hour for an individual familiar with the method. Navy Method This method, which is based largely upon a radex wind plot, is a system of weighing incremental square areas according to the degree of fall-out expected, relative to one another. 52 It results