CONFIDENTIAL 3. Radial roads to permit mass evacuation of majorcities could not be built in time for this tactic to have any value. 4. Thefeasibility of an underground shelter construction program has not been proven to the point of stimulating Congress to provide public funds. The technical know-how ne ai ee for large shelter construction exists, although one of the most promising and inexpensive designs has not been tested. Effectiveness 5. In the Washington area, against attack with one to four 10-Mt ground-burst weapons all aimed at the population center with a 4000-m cep, results were as follows: a. Use of underground shelter, evacuation with 1959 warning time, and evacuation with 1956 warning time are all more effective civil defense measures than use of existing shelter. b. Use of underground shelter and evacuation with 1959 warning time are more effective than evacuation with 1956 warning time. c. Use of underground shelter is more effective against several weapons than evacuation with 1959 warning time, and is as effective as such evacuation against a single weapon. 6. In the Boston area, against attack with 10-Mt ground-burst weapons aimed at the population center, use of underground shelter is more effective than any othercivil defense measure for all weights of attack from one to four weapons and for all cep from 4000 to 12,000 m, even wheneffects of fallout are completely ignored; the superiority of underground shelter is further increased when falloutis considered. 7, In the Milwaukee and St. Louis areas, against attack with 10-Mt ground-burst weapons, wheneffects of fallout are ignored, evacuation with 1959 warning time is the most effective measure for a cEP of 4000 m, and undergroundshelteris the most effective measure for a cep of 12,000 m. Whenfallout is considered, the superiority of evacuation with 1959 warning time for a cep of 4000 m is reduced and perhapseliminated, depending on thelocal and regionalfallout pattern, and the superiority of underground shelter for a cEP of 12,000 m is further increased. 8. In the Dayton area, against attack with 10-Mt ground-burst weapons, wheneffects of fallout are ignored, evacuation with 1959 warning time is the most effective measure for all cep from 4000 to 12,000 m. Whenfallout is considered, the superiority of evacuation is reduced and perhaps eliminated, depending on the local and regional fallout pattern. 9. Any increase in radiation effects resulting from attacks on other nearby targets will increase the effectiveness of undergroundshelter relative to the other possible civil defense tactics. This relative superiority will be most drastic when the total fallout intensity reaches a level where the 0.9 protection factor of best shelter now available permits occupants to receive a lethal dose. 10. Shelter that will attenuate radiation effects by 0.9 (ordinary basement shelter) ix not adequate in urban targets: a. At 2 to 4 miles from ground zero, individuals in basement shelters would receive an LD, dose in 3 hr; at 4 to 5 miles, in 6 hr. At these distances fallen trees and other debris in a high radiation field would make rescue operations impossible within the hours of life left to occupants of basement shelters. b. At distances that might be relatively debris-free (7 to 8 miles), LD». doses would be received by occupants of basement shelters after 24 hr. Evacuation by shielded vehicle would be imperative to preservelife. A ORO-R-17 (App B) CONFIDENTIAL