Reproducibility of surface wind and tracer transport simulations over
complex terrain using 5, 3, and 1 km grid models
Abstract
It was investigated that the reproducibility of surface wind and tracer
transport simulations over complex terrain in wintertime using
high-resolution (5-km, 3-km, and 1-km grid) weather and transport
models, in which radioactive cesium (Cs-137) emitted from the Fukushima
nuclear power plant was used as a tracer. Fukushima has complex terrain,
such as mountains and valleys. The model results were validated by
observations collected from the national networks of the automated
meteorological data acquisition system and the hourly air pollution
sampling system. The reproducibility depended on the model resolution,
topographic complexity, and synoptic weather conditions. Higher model
resolution led to higher reproducibility of surface winds, especially in
mountainous areas when the Siberian winter monsoon was disturbed. In
contrast, the model improvement was negligible or nonexistent over
plain/coastal areas when the synoptic field was steady. The statistical
scores of the tracer transport simulations often deteriorated due to
small errors in the plume locations. However, the higher-resolution
models advantageously performed better transport simulations in the
mountainous areas because of the lower numerical diffusion and higher
reproducibility of the mass flux. The reproducibility of the tracer
distribution in the valley of the Fukushima mountainous region was
dramatically improved with increasing model resolution. In conclusion, a
higher-resolution model is definitely recommended for tracer transport
simulations over mountainous terrain at least in the range of mesoscale
model resolutions (commonly 1~10 km grids).