Fallout sheltering: is it feasible?

Abstract

The feasibility of sheltering the U.S. population from fallout resulting from a large-scale nuclear attack is assessed using a mathematical model. The model is used to calculate the reduction in cumulative dose received by a sheltered survivor, as a function of five adjustable parameters. Three time periods are postulated: time in the shelter, a transition period during which time out of the shelter increases and a final period in which half the time is spent outside the shelter. The parameters are varied independently, and the resulting dose reduction factor is compared with what seems to be necessary for survival in different regions of the country under the postulated attack. Another model developed by K.S. Gant and C.V. Chester is compared with this one. Similarities and differences are pointed out, and where possible the results of the two are checked for compatibility. An important question addressed in this paper is whether under the conditions of a large-scale nuclear attack sheltering a relatively unprepared population is at all feasible. Sensitivity tests of the various parameters in our model show that relatively low protection factor areas such as basements or inner rooms already existing in homes or other buildings could quite adequately serve as shelters for most of the area of the contiguous United States. Furthermore, continuous stays in these shelters of more than three weeks do not seem to be necessary for these large parts of the United States.