講演情報
[AAS08-P03]High resolution CO2 simulation around the Takayama forest in central Japan
*長谷川 朝香1,2、今須 良一1、佐藤 陽祐3、Wang Qiao1、新井 豊1、村山 昌平2,4、近藤 裕昭2、石戸谷 重之2、前田 高尚2 (1.東京大学大気海洋研究所、2.国立研究開発法人産業技術総合研究所、3.大阪大学大学院工学研究科 、4.獨協大学)
キーワード：
二酸化炭素、CO2フラックス、大気輸送モデル、陸域生態系モデル
Achieving carbon neutrality requires an objective evaluation method of decarbonization technologies and policies. Inverse analysis enables the estimation of CO2 emissions and uptake by combining an atmospheric transport model with observations. Because this approach relies on observational data, it provides a more objective estimate of CO2 inventory compared with bottom-up inventories. According to IPCC (2022), CO2 emissions from urban areas represent approximately 70% of global emissions. A high-resolution atmospheric transport model, capable of simulating at 1km or finer resolution, is required to estimate CO2 transport and flux within urban areas. When estimating CO2 fluxes in metropolitan areas such as Tokyo metropolitan region, it is necessary to consider not only anthropogenic emissions but also biogenic fluxes from suburban forests. To estimate anthropogenic CO2 fluxes in regions where various emission and uptake sources coexist, it is essential to first understand CO2 fluxes and concentration variations over the area where vegetation is the dominant source of CO2 exchange. In this study, we focused on biogenic CO2 fluxes and simulated CO2 concentration distributions around the forest using a terrestrial ecosystem model and a high-resolution atmospheric transport model. Few studies have evaluated CO2 distributions around forested areas in Japan using these high-resolution models.
Our target site is Takayama (TKY) in Gifu Prefecture, located in central Japan. At this site, CO2 fluxes, CO2 concentrations, and meteorological variables are observed at and around a flux tower (Murayama et al., 2024). We calculated hourly CO2 concentration variations during the growing season with SCALE-RM-CO2 (Sato et al., 2025). The calculation resolutions range from 10 km to 1 km using offline nesting. Both anthropogenic and natural fluxes were prepared as at a 1 km resolution, based on EAGrid 2000-Japan (Kannari et al., 2007) and the terrestrial ecosystem model, BEAMS-diurnal (Wang et al., 2021), respectively. This 1 km fluxes dataset was aggregated to match the resolution of each transport calculation. We first evaluated the performance of BEAMS-diurnal in reproducing observed CO2 fluxes at this site. BEAMS-diurnal captures the diurnal variation of gross primary production satisfactorily, but its amplitudes of ecosystem respiration are larger than those in the observations. SCALE-RM-CO2 reproduced the seasonal and diurnal variations, as well as the monthly mean CO2 concentrations, reasonably well at this site. The correlation between the model results and observations was about 0.6 for 1 km resolution for August 2016. In the presentation, we will also show the comparison results between the models and observations for other months, as well as analyses of local CO2 transport considering the complex terrain around this site.
Our target site is Takayama (TKY) in Gifu Prefecture, located in central Japan. At this site, CO2 fluxes, CO2 concentrations, and meteorological variables are observed at and around a flux tower (Murayama et al., 2024). We calculated hourly CO2 concentration variations during the growing season with SCALE-RM-CO2 (Sato et al., 2025). The calculation resolutions range from 10 km to 1 km using offline nesting. Both anthropogenic and natural fluxes were prepared as at a 1 km resolution, based on EAGrid 2000-Japan (Kannari et al., 2007) and the terrestrial ecosystem model, BEAMS-diurnal (Wang et al., 2021), respectively. This 1 km fluxes dataset was aggregated to match the resolution of each transport calculation. We first evaluated the performance of BEAMS-diurnal in reproducing observed CO2 fluxes at this site. BEAMS-diurnal captures the diurnal variation of gross primary production satisfactorily, but its amplitudes of ecosystem respiration are larger than those in the observations. SCALE-RM-CO2 reproduced the seasonal and diurnal variations, as well as the monthly mean CO2 concentrations, reasonably well at this site. The correlation between the model results and observations was about 0.6 for 1 km resolution for August 2016. In the presentation, we will also show the comparison results between the models and observations for other months, as well as analyses of local CO2 transport considering the complex terrain around this site.