environment, the shielding circumstances of the survivors, and the body shielding of the various organs. It is a modular system with separate databases for each of the free-field radiation components, for each of several distinct shielding environments, and for each of many different organs.
The free-field components include prompt neutrons, early gamma rays (prompt-fission gamma rays and gamma rays from inelastic scattering and capture of prompt neutrons), late gamma rays (from fission products and from delayed neutrons), and delayed neutrons. A new or revised treatment of any of those components can readily be introduced by appropriately substituting a new database for an existing one. The shielding databases include models for all survivors with nine-parameter shielding and all survivors with globe-data shielding descriptions (Roesch 1987).
Uncertainties were estimated for the shielding and organ environments by calculating fractional standard deviations among the many shielding and phantom environments that had been computed. They were combined with uncertainties in the free-field radiation fluences to provide a preliminary estimate of uncertainty in the computed doses. Uncertainty evaluation was incomplete at the time of the adoption of DS86 (Roesch 1987). Later, Kaul and Egbert (1989) presented the US dosimetry committee with a draft of a preliminary uncertainty analysis; this analysis was revised in 1992 but is still regarded as preliminary.
In the DS86 calculations of kerma and dose to exposed people, the codes and data used were superior to any used previously. Especially for doses to organs deep in the body, gamma rays dominate, amounting in Hiroshima to 98–99% of the total absorbed dose. In Nagasaki, where the neutron fluence at a specified total dose is only about one-third that in Hiroshima, the percentage contribution of neutrons is even lower.
Not only were the calculations of gamma rays believed to be improved in DS86, but a most important consideration was the experimental confirmation of gamma-ray doses by measurements (with thermoluminescent dosimetry or TLD) of the gamma-ray signal in quartz, brick, and tile samples in both cities. Agreement between measurement and calculation is quite good over a wide range of distances from the hypocenter in both cities. That bears repeating for emphasis: the most important component of the dose, gamma rays, is experimentally well verified (see Chapter 2).
The contribution of fast neutrons to the dose in organs deep in the body is estimated in DS86 at around 1–2% of the total absorbed dose in Hiroshima in the dose range of about 0.5–2 Gy. Nevertheless, especially because of the potentially