[MAG33-P09] impact of observed precipitation on the simulation of
diffusion/deposition process of the radioactive nuclides
from the Fukushima Daiichi Nuclear Power Plant
On March 11, 2011, Tohoku Region Pacific Coast Earthquake occurred, and the resulting tsunami hit the Fukushima Daiichi nuclear power plant. Large amounts of radioactive materials were discharged into the atmosphere, transported, and deposited on the surface. Meteorological variables such as pressure, temperature, wind, humidity, and precipitation, are actively involved in the transport and deposition processes of the radioactive materials. Atmospheric Transport Models (ATMs) are used to predict their behavior. However, uncertainty remains in the predicted meteorological variables, especially regarding the precipitation process. In this study, in order to reduce uncertainty in the prediction of the precipitation process, we compared two sets of simulations: One with normal run, the other with the same but only compare the output precipitation was replaced by an observation.
In this study, we used an ATM called the Lagrangian Model (LM; Kajino et al., 2016). It has a feature that it is light and easy to run for non-specialists of modeling research who may not have large computational resources, or for the computationally expensive analysis such as ensemble forecasts and sensitivity simulations. For the meteorological variables as input of LM, we used the Japan Meteorological Agency's Grid Point Value Meso Scale Model (MSM). This data is obtained every 3 hours. Data such as pressure, geopotential height, temperature, humidity, wind, and precipitation amount are used as the input. Among them, only precipitation was replaced with an observation data. The Radar - Automated Meteorological Data Acquisition System (RA) was used for the observation data. Katata et al. (2015) was used for the emission scenario. The simulation data was compared with an observation data of surface air activity concentration of 137Cs. The observation data was obtained from Suspended Particulate Matter (SPM) tapes by Oura et al. (2015). SPM refers to an aerosol having a diameter of smaller than 10 μm.
The simulation result was better with RA than the normal run in terms of the comparisons against the actual measurement values of 137Cs surface activity concentrations. Since the time interval of MSM outputs was every 3 hours, it is difficult to reproduce precipitation events shorter than 3 hour. On the other hand, since RA outputs has a shorter interval, 30 minutes, it could reproduce details of the precipitation process more such as cloud breaks and sudden precipitation. The difference of the time interval caused difference in simulated wet deposition amount and a surface concentrations. Therefore, in the ATM, it was found that the distribution and the timing of precipitation process were more important, than the amount of precipitation. Through this study, the importance of the precipitation process in the ATM for the 137Cs simulation was reconfirmed.
In this study, we used an ATM called the Lagrangian Model (LM; Kajino et al., 2016). It has a feature that it is light and easy to run for non-specialists of modeling research who may not have large computational resources, or for the computationally expensive analysis such as ensemble forecasts and sensitivity simulations. For the meteorological variables as input of LM, we used the Japan Meteorological Agency's Grid Point Value Meso Scale Model (MSM). This data is obtained every 3 hours. Data such as pressure, geopotential height, temperature, humidity, wind, and precipitation amount are used as the input. Among them, only precipitation was replaced with an observation data. The Radar - Automated Meteorological Data Acquisition System (RA) was used for the observation data. Katata et al. (2015) was used for the emission scenario. The simulation data was compared with an observation data of surface air activity concentration of 137Cs. The observation data was obtained from Suspended Particulate Matter (SPM) tapes by Oura et al. (2015). SPM refers to an aerosol having a diameter of smaller than 10 μm.
The simulation result was better with RA than the normal run in terms of the comparisons against the actual measurement values of 137Cs surface activity concentrations. Since the time interval of MSM outputs was every 3 hours, it is difficult to reproduce precipitation events shorter than 3 hour. On the other hand, since RA outputs has a shorter interval, 30 minutes, it could reproduce details of the precipitation process more such as cloud breaks and sudden precipitation. The difference of the time interval caused difference in simulated wet deposition amount and a surface concentrations. Therefore, in the ATM, it was found that the distribution and the timing of precipitation process were more important, than the amount of precipitation. Through this study, the importance of the precipitation process in the ATM for the 137Cs simulation was reconfirmed.