Atmospheric resuspension of radionuclides can be a secondary source of cont
amination after a release has stopped, as well as a source of contamination
for people and areas not exposed to the original release. A test scenario
based on measurements collected after the Chernobyl accident was used to ev
aluate existing mathematical models for contaminant resuspension from soil,
to examine resuspension processes on both local and regional scales, and t
o investigate the importance of seasonal variations of these processes. Mod
el predictions from 15 participants were compared with measured air concent
rations and resuspension factors to investigate and explain the discrepanci
es both among model predictions and between model predictions and observati
ons and thus to evaluate the predictive capabilities and drawbacks of commo
nly used generic resuspension models. The empirical models tested can give
predictions within an order of magnitude of observations or better if adequ
ately calibrated for site-specific conditions, but they do not describe the
process-level events or account for spatial heterogeneity or temporal vari
ations. (C) 1998 Elsevier Science Ltd. All rights reserved.