High-resolution comparison of coupled atmospheric dispersion systems for a hypothetical I-131 release.

Abstract

High-resolution atmospheric transport and dispersion predictions may diverge when different meteorological and dispersion models are coupled. We compared two coupled meteorological-dispersion systems, HYSPLIT-HARMONIE and FLEXPART-WRF, for a hypothetical 24 h I-131 release from Heysham nuclear power station. Four paired simulations combined April and October 2015 meteorological cases with gaseous and particulate I-131. Source timing, activity, release profile, radioactive decay, output grid, vertical output layers, and fixed dry-deposition velocities were aligned. Spatial agreement was evaluated using area-weighted figure of merit in space (FMS) for ground-level and total-plume threshold-exceedance areas at a study-defined 250 Bq m-3 threshold, with sensitivity analyses at 50 and 1000 Bq m-3. At 20 h, ground-level FMS was 14.39-19.49%, and total-plume FMS was 27.09-33.55%. October ambient activities differed by 1.96-2.81% despite ground-level FMS below 15.00%. At 20 h in April, the systems had similar within-domain sums of ambient and deposited activity, but the separate components and predicted plume locations differed. The comparison shows why plume location should be reported alongside integrated activity when multiple modelling systems inform emergency monitoring.