4:30 PM - 4:45 PM
[MAG33-09] Effects of aerosols derived from fires after the atomic bomb explosion on rainfall distribution in Hiroshima
Keywords:aerosols, precipitation, regional model, atmospheric chemistry
Aerosols in the atmosphere reflect or absorb radiation and modulate cloud formation and characteristics. Black carbon particles, a part of aerosols, can also absorb solar radiation and heat the surrounding atmosphere, thereby altering the vertical distribution of atmospheric temperature, which can modulate atmospheric stability and precipitation. These processes are being implemented into recent climate models as potentially significant influences on climate change, and negligible effects might also be expected after sporadic releases of aerosols such as biomass burning.
We have conducted numerical simulations using a regional chemical transport model to evaluate the effect of the interaction between aerosols emitted from fires and cloud/precipitation processes. The emission scenario of aerosols and latent and sensible heat was taken from previous studies and recent observations of forest fires. In addition, the impact of uncertainties in the meteorological fields on the downscaling calculations around Hiroshima was estimated by conducting ensemble experiments with ten members of the ECMWF ERA5 EDA (ensembles of data assimilations).
Comparison with the observations of ground meteorological elements at the Hiroshima Meteorological Observatory, digitized by the Hirosaki University group, showed good reproducibility for temperature and wind direction. The temperatures tended to be overestimated around noon on August 6, 1945, when fires were most active. Still, fires may have influenced this in the surrounding area since fires were also reported in the Eba district at the foot of the observatory. After the fires broke out in the city center, where strong upwind currents were generated, the winds predominantly blew toward the city center, and all members generally indicated a northward wind. These results suggest that improving the reliability of the spatiotemporal distribution of fires is critical to improving the reproducibility of wind direction near the ground surface.
As for the rainfall distribution, no rainfall was observed around the city center when there were no fires in Hiroshima, but it was estimated that there was up to 200±40 mm of accumulated rainfall near the hypocenter until 18:00 on August 6, when the fires were approximately extinguished in the case with the fires without considering the interaction. The difference in precipitation among the members is relatively large west of the hypocenter, suggesting that the wind direction uncertainty in the lower free troposphere affected the rainfall distribution in the area. In the "with interaction" case, precipitation near the hypocenter decreased by more than 100 mm compared to the "without interaction" case. In comparison, it increased by up to 5 mm over a wide area in the surrounding region. It might be caused by a semi-direct effect of aerosols-enhanced atmospheric heating in the fire area around the hypocenter, increasing the boundary layer altitude by nearly 1 km, from 2.7 ± 0.4 km to 3.8 ± 0.4 km, and increasing the air temperature near the surface by 5 ± 1.2 K and in the upper troposphere at an altitude of 8-9 km by about 2 ± 1 K, The enhanced upwelling and increased saturated vapor pressure may also have contributed to the reduced occurrence of rainfall in the vicinity of the hypocenter. The structure of the horseshoe-shaped rainfall area around the hypocenter, pointed out by Masuda (1989), was reproduced only in the "with interaction" case. It suggests that aerosol effects are not negligible for estimating the distribution of "black rain" especially near the fire area.
We have conducted numerical simulations using a regional chemical transport model to evaluate the effect of the interaction between aerosols emitted from fires and cloud/precipitation processes. The emission scenario of aerosols and latent and sensible heat was taken from previous studies and recent observations of forest fires. In addition, the impact of uncertainties in the meteorological fields on the downscaling calculations around Hiroshima was estimated by conducting ensemble experiments with ten members of the ECMWF ERA5 EDA (ensembles of data assimilations).
Comparison with the observations of ground meteorological elements at the Hiroshima Meteorological Observatory, digitized by the Hirosaki University group, showed good reproducibility for temperature and wind direction. The temperatures tended to be overestimated around noon on August 6, 1945, when fires were most active. Still, fires may have influenced this in the surrounding area since fires were also reported in the Eba district at the foot of the observatory. After the fires broke out in the city center, where strong upwind currents were generated, the winds predominantly blew toward the city center, and all members generally indicated a northward wind. These results suggest that improving the reliability of the spatiotemporal distribution of fires is critical to improving the reproducibility of wind direction near the ground surface.
As for the rainfall distribution, no rainfall was observed around the city center when there were no fires in Hiroshima, but it was estimated that there was up to 200±40 mm of accumulated rainfall near the hypocenter until 18:00 on August 6, when the fires were approximately extinguished in the case with the fires without considering the interaction. The difference in precipitation among the members is relatively large west of the hypocenter, suggesting that the wind direction uncertainty in the lower free troposphere affected the rainfall distribution in the area. In the "with interaction" case, precipitation near the hypocenter decreased by more than 100 mm compared to the "without interaction" case. In comparison, it increased by up to 5 mm over a wide area in the surrounding region. It might be caused by a semi-direct effect of aerosols-enhanced atmospheric heating in the fire area around the hypocenter, increasing the boundary layer altitude by nearly 1 km, from 2.7 ± 0.4 km to 3.8 ± 0.4 km, and increasing the air temperature near the surface by 5 ± 1.2 K and in the upper troposphere at an altitude of 8-9 km by about 2 ± 1 K, The enhanced upwelling and increased saturated vapor pressure may also have contributed to the reduced occurrence of rainfall in the vicinity of the hypocenter. The structure of the horseshoe-shaped rainfall area around the hypocenter, pointed out by Masuda (1989), was reproduced only in the "with interaction" case. It suggests that aerosol effects are not negligible for estimating the distribution of "black rain" especially near the fire area.