3:00 PM - 3:15 PM
[AAS02-06] Effects of spatio-temporally-varying anthropogenic heat on high-resolution modelled global climate
Keywords:Global climate modeling, Anthropogenic heat emission, Global urban climatology
Because of energy consumption, urban areas emit human-induced heat to the atmospheric environment. This so-called anthropogenic heat emission (AHE) is an additional heat source from the surface to the atmosphere above. In highly dense cities, temporally-varying AHE is a significant heat source, in addition to the solar radiation and long-wave radiation forcing. Hence, multiple numerical studies have been conducted to investigate its regional atmospheric effects. Meanwhile, global datasets of hourly-varying AHE have become more available at high spatial resolutions (e.g. 1-km). Despite this, the global effects of AHE are not well understood. To address this, this work investigated the global climate modifications attributed to the AHE by using a 14-km resolving global climate model.
The Nonhydrostatic ICosahedral Atmospheric Model (NICAM), a global cloud resolving model, was modified to be able to read spatially-representative AHE input (i.e. AH4GUC). Here, hourly-changing AHE typical for a given month was also considered by updating its value depending on the modelled actual time in UTC. Inside the model, AHE was assumed to be an added increment of the calculated sensible heat flux. This can be implemented under two possible cases: (1: ahematsiro) adding AHE only to the calculated sensible heat flux within the land surface model (i.e. Matsiro model); or (2: ahesfc) adding AHE to the grid-total sensible heat flux calculated by the surface mode. This process is done for each grid at each time step.
After model modifications, simulations were conducted for July 2023, a year recorded to be extremely hot according to NASA. Comparisons of the ahematsiro and ahesfc with a default case (3: noahe) that does not consider AHE was conducted. Full-month simulations initialized at 00UTC July 1, 2023 with ERA-5 reanalyses were done for all three cases. Furthermore, multiple time-lagged ensembles were conducted for 36 hours for ahematsiro and ahesfc by branching off from the full-month noahe simulation at 12-hour intervals from 00UTC July 15, 2023. All cases were successfully compiled and smoothly modelled using the Tsubame 3.0 supercomputer of Tokyo Institute of Technology. Apart from the AHE setup, default settings of NICAM Ver. 21 were used throughout.
Directly influenced by surface heat flux, near-surface (2-m above ground) air temperatures (T2) were initially evaluated. For validation, conservative regridding was done for the simulations and ERA-5 reanalyses at 1.0 degrees spacing. The zonal average of the full-month simulated T2 follows a similar pattern as that from the ERA-5 reanalysis. Spatially, all cases underestimated the mean T2 over the ocean in the Southern Hemisphere, with increased differences towards the poles.
Unlike limited-area models (e.g. downscaling model) where the internal domain's simulated climate does not influence the climate outside the lateral boundaries, NICAM reveals the global climate modification by AHE. The hourly time-series of global mean T2 reveals that the simulated climate of the AHE cases diverges significantly from the noahe case after 7 days since initialization, while maintaining the diurnal amplitude of global temperatures throughout the simulation time. After a week has been simulated, large-scale circulations or the global climate itself has been modified. Meanwhile, the ensembles reveal that the influence of AHE to temperature begins locally and widens over the country more than a day of simulation. In large cities, the correlation of its AHE to its local T2 increase (AHE cases minus noahe) is significant only for the first 3-days, after which the AHE-induced warming influences large-scale dynamics. A deeper ensemble analysis is underway to extract highly probable effects of AHE to global climate not only on the surface but aloft.
This work is supported by JHPCN (JH230052), the Scientific Research (A) 21H04573, and Grant-in-Aid for Early Career Scientists 21K14249 of MEXT (Japan).
The Nonhydrostatic ICosahedral Atmospheric Model (NICAM), a global cloud resolving model, was modified to be able to read spatially-representative AHE input (i.e. AH4GUC). Here, hourly-changing AHE typical for a given month was also considered by updating its value depending on the modelled actual time in UTC. Inside the model, AHE was assumed to be an added increment of the calculated sensible heat flux. This can be implemented under two possible cases: (1: ahematsiro) adding AHE only to the calculated sensible heat flux within the land surface model (i.e. Matsiro model); or (2: ahesfc) adding AHE to the grid-total sensible heat flux calculated by the surface mode. This process is done for each grid at each time step.
After model modifications, simulations were conducted for July 2023, a year recorded to be extremely hot according to NASA. Comparisons of the ahematsiro and ahesfc with a default case (3: noahe) that does not consider AHE was conducted. Full-month simulations initialized at 00UTC July 1, 2023 with ERA-5 reanalyses were done for all three cases. Furthermore, multiple time-lagged ensembles were conducted for 36 hours for ahematsiro and ahesfc by branching off from the full-month noahe simulation at 12-hour intervals from 00UTC July 15, 2023. All cases were successfully compiled and smoothly modelled using the Tsubame 3.0 supercomputer of Tokyo Institute of Technology. Apart from the AHE setup, default settings of NICAM Ver. 21 were used throughout.
Directly influenced by surface heat flux, near-surface (2-m above ground) air temperatures (T2) were initially evaluated. For validation, conservative regridding was done for the simulations and ERA-5 reanalyses at 1.0 degrees spacing. The zonal average of the full-month simulated T2 follows a similar pattern as that from the ERA-5 reanalysis. Spatially, all cases underestimated the mean T2 over the ocean in the Southern Hemisphere, with increased differences towards the poles.
Unlike limited-area models (e.g. downscaling model) where the internal domain's simulated climate does not influence the climate outside the lateral boundaries, NICAM reveals the global climate modification by AHE. The hourly time-series of global mean T2 reveals that the simulated climate of the AHE cases diverges significantly from the noahe case after 7 days since initialization, while maintaining the diurnal amplitude of global temperatures throughout the simulation time. After a week has been simulated, large-scale circulations or the global climate itself has been modified. Meanwhile, the ensembles reveal that the influence of AHE to temperature begins locally and widens over the country more than a day of simulation. In large cities, the correlation of its AHE to its local T2 increase (AHE cases minus noahe) is significant only for the first 3-days, after which the AHE-induced warming influences large-scale dynamics. A deeper ensemble analysis is underway to extract highly probable effects of AHE to global climate not only on the surface but aloft.
This work is supported by JHPCN (JH230052), the Scientific Research (A) 21H04573, and Grant-in-Aid for Early Career Scientists 21K14249 of MEXT (Japan).