INTRODUCTION The body of this appendix indicates that the most effective method of minimizing urban population casualties under thermonuclear attack is the use of underground shelters providing a high level of resistance to peak blast overpressure and adequate shielding against initial and residual radiation. The long-range inferiority of alternative passive defense measures is due to some special characteristic of each. Best existing shelter offers very poor protection against blast and would supply only a few hours of life under local fallout from megaton weapons. Mass evacuation is time-consuming at best and may result in panic at worst; in the case of missile attacks this measure cannot be undertaken, owing to short warning time. Under most conditions of airborne attack, use of underground shelters would yield more survivors than use of best existing shelter or evacuation. The relative superiority of underground shelter increases as the number of weapons per target increases and as aiming accuracy deteriorates. Furthermore undergroundshelter is undoubtedly the best passive defense measure under the conditions of short warning time associated with missiles, whether submarine-launched or IcBM, and against the widespread high-intensity fallout from a mass attack. LIMITATIONS OF UNDERGROUND SHELTER To be usable under the short warning of missile attack, underground shelter must be located close to where people are, both by day and by night. Construction of a system of underground shelters on a geographical pattern matching the present distribution of daytime and nighttime populations would result in a concentration of shelters itself vulnerable to the cratering effects of megaton weapons and the very high blast pressures just beyond the crater and lip. If the accuracy of enemy bombing were poor, some concentrations of shelters would escape; however, the possibility of a high cer cannot be relied on to remove the hazard of direct hits on shelter concentrations, and cannot be established by national policy as a passive defense measure. Thus the prime limitation of an underground shelter program is the necessity to spread out the pattern of shelter sites and at the same time keep people close to their shelters. The second major limitation is cost. Previously proposed shelter programs, such as those recommended to Congress by rcpa during past years, have been summarily rejected by Congressional committees on the grounds of cost. The possibilities of favorable Con. gressional action are improving and it is entirely possible that the cost of a national program of shelters able to withstand 30 to 100 psi and giving adequate radiation attenuation may soon be acceptable to Congress. The concentration of shelters in potential crater areas may create a new stumbling block. This limitation might be ameliorated by sinking shelters so deep through use of shafts and tunnels that they would be below any surface-burst crater or lethal underground shock. But again, Congressional appropriations leaders might shy at the cost of this shelter program, which would obviously be much higher than that for shelters just below ground level. ORO-R-17 (App B) 83 CONEIRENTIAL.