260 likes | 457 Views
Typhoon Simulation by Using a Global Cloud Resolving Model on Earth Simulator. development of the model preliminary result of typhoon simulation. W. Yanase, S. Iga, T. Nasuno, H. Miura, H. Tomita, and M. Satoh 31 st October, 2006. Targets of Global Cloud-Resolving Model.
E N D
Typhoon Simulation by Using a Global Cloud Resolving Model on Earth Simulator • development of the model • preliminary result of typhoon simulation W. Yanase, S. Iga, T. Nasuno, H. Miura, H. Tomita, and M. Satoh 31st October, 2006
Targets of Global Cloud-Resolving Model • Multi-scale convection • Tropical cyclones • Madden Julian Oscillation • Cloud clusters • Effects of cumulus clouds on climate • condensational heating • precipitation • vertical transport • radiation
Development of Global Model “NICAM” “Nonhydrostatic” + “ICosahedral” Atmospheric Model Icosahedral Glevel-0 Glevel-1 Glevel-3 Glevel-5 current dx = 14 km Glevel-9 dx = 7 km Glevel-10 dx = 3.5 km Glevel-11
Schemes of Physics in NICAM • Turbulance: Mellor & Yamada level-2 • Radiation: MSTRNX (Sekiguchi & Nakajima, 2006) • Cloud microphysics: Grabowski (1998) • Cumulus convection • Arakawa & Shubert for dx > 30 km • Not used for dx = 3.5 km, 7 km, 14 km
History of NICAM • First simulation (e.g. Tomita & Satoh, 2004) • Aqua planet experiment (e.g. Tomita et al., 2005) • dx = 3.5 km, 10-day integration • eastward propagating multi-scale clusters • Real topography (e.g. Miura et al.; submitted to GRL) • dx = 3.5 km, 7 km, 14 km • simulation of a typhoon in Apr. 2004 • Simulations with dx=3.5km, 7km, 14km are performed on Earth Simulator computer, and the results are currently analyzed by scientists in Japan & USA
Experimental Design of Typhoon Simulation • Model: NICAM • Topography: GTOPO30 (smoothed) • Initial condition: • NCEP tropospheric analyses (1.0deg x 1.0deg) • 00:00UTC on 1st April 2004 • no bogusing modifications • Time integration • 7 days for dx = 3.5 km (Glevel-11) • 10 days for dx = 7 km (Glevel-10) • 30 days for dx = 14 km (Glevel-9)
Satellite Observation & Model Results OLR Apr 02 00UTC dx~3.5 km dx~7 km dx~14 km Apr 03 00UTC (http://weather.is.kochi-u.ac.jp/)
Satellite Observation & Model Results Apr 04 00UTC dx~3.5 km dx~7 km dx~14 km Apr 05 00UTC
Satellite Observation & Model Results Apr 06 00UTC dx~3.5 km dx~7 km dx~14 km Apr 07 00UTC
Precipitation (April 5th) AMSR-E dx~3.5 km dx~7 km dx~14 km
Radial-Vertical Structure (dx=14km: Apr. 7th 12UTC) tangential wind z=10km radial wind vertical wind r=500km
Radial-Vertical Structure (dx=14km, Apr. 7th 12UTC) temperature anomaly z=10km relative humidity condensed water r=500km
Future Plans • Technical issue • improvement of turbulence scheme • Kain-Fritsch scheme for dx > 7 km • Case study of TC: dx = 3.5 km, ~10 day • validation with observational data • Formation of TC: dx = 7 km, 14 km, ~30 day • little influence of initial condition • Climatology of TC: dx = 30 km , 60km; 1 year ~ • distribution of genesis and development
Computational Costs • SR11000(20proc): 100-day real time • glevel-6 (120km): ~20000day • glevel-7 (60km):~2500day • glevel-8 (30km): ~300day • Earth Simulator: 100,000 node-hour a year • glevel-8 (30km): ~10000day • glevel-9 (14km): ~1200day • glevel-10 (7km): ~300day • glevel-11 (3.5km): ~40day
Radial-Vertical Structure (dx=14km) water vapor z=10km potential temperature equivalent P.T. r=500km
Radial-Vertical Structure (dx=14km) cloud water z=10km rain snow r=500km
Zonal-Vertical Section: dx=14km meridional wind vertical wind
Parameter of Turbulence (mean over 10S-10N)