ICRP model for 137Cs referred to in the text (Figure D1). For large n; and r tj distributed randomly throughout each year, it follows that total int ted whole-body dose Q(#) in Bq kg" after t years may be approximated by the quantity hs , ra\cbas} «rex, where X is here defined as the braced quantity and where the variafe T, subsumed in S, is here—in contrast to Eq. (2) above—uniformly distriquted between 0 and?. Based on Eq. (4) and the preceding analysis, interindividual vari ility in expected dose (D(t)) by time t was characterized by evaluating (D(t)) = [(0.36 + Y)D,(#) 1+ D,,(t) + (F) {et(R){S)} , A6) in which Y was defined in the text and (S) , the expectation of S with respect to both T and 8,is given by iS)=1+ AB +e“[Ei(b,) -Ei(b,)]-Ei(c,)+Ei(c,)+Ln(c,/ cy) ABKAt b, =-BKt, i=0,1, C; =b.-At, i=0,1, and AB = (B, — By) = (1.107 — 0.9) = 0.207 ‘in which Ei is the exponential integral. As such, variability in (D(t)) from uniform variability in F and lognormal variability in both (R) ises nd H (see text). Uncertainty in population-average dose D(t) was characteriged by evaluating (A7) DW =D,@)+D,()+ FBX , D-5