Problems in using this procedure have been particularly difficult very close to ground zero and very far away from it. At close-in locations, physical factors have often prevented the installation of recording instruments, and high radiation levels have denied entry for standard methods of measurement until decay and weathering have greatly detreased the levels of radiation. Airplaneand helicopter measurements over such areas have not been reliable. Often, however, because the area within the innermost measured contour is small, the resultant K-factor has not been sensitive to the estimates needed in lieu of measurements within that contour. At great distances, the reliability of measurements is reduced because intensities are small, approaching background levels. Unfortunately their contribution to the integral can be large because of the large areas involved. Properly, A; (or rather the value of 1(A})) should be determined by the definition of local fallout. The tendency of many investigators to carry out the integration to the limit of reliability of the data results in an implicit definition of local fallout that varies from shot to shot and makes intercomparison of results difficult. A number of empirical values of Ky are listed in Table 1. (We cannot guarantee that these data do meet the criterion of consistent integration limits.) All the fallout patterns from which these data were obtained are uncertain to some degree. The Subcommittee believes / that the best near-surface-burst data on this list are those from Johnie Boy, Buffalo 2, Zuni, Tewa, and Jangle Surface. Taking a mean of those average values, we get Ky » 1090. This mean represents average field-roughness conditions, and instruments as used in the past. In DCPA use, a K-factor is required which does not include corrections for surface roughness or instrument response, which is to say Ky. Since Kj = 16/9 K,, the result is K) a 1930, The Subcommittee recommends that DCPA use this value of K}t A