Hello,
I have the model setup with 3 nested domains with inner grid size of 700x700 with 5 km resolution. Other domains are in the multiples of 3. I have to estimate the Model processing and storage requirement for 3 days of forecast in 3 hr interval with GFS 0.5 deg data set. The other model configuration is listed below.
#Domain
Time_step = 75 sec
e_vert = 30
#Phyics
sf_ocean_physics = 0
chem_opt = 0
mp_physics = 3, WSM 3-class simple ice scheme
ra_lw_physics = 1, rrtm scheme
ra_sw_physics = 1, Dudhia scheme
radt = 50 min
sf_sfcaly_physics = 1, MM5 Monin-Obukhov scheme
sf_surface_physics = 2, Unified Noah land-surface model
sf_urban_physics = 0, activate urban canopy model
bl_pbl_physics = 1, YSU scheme
bldt = 0
cu_physics = 1, Kain-Fritsch scheme
isfflx =1
ifsnow = 0
icloud = 1
#dynamics
rk_ord = 3
Dear Amit,
As far as I know, the only way to measure storage capacity and simulation integration time is to run a test simulation.
For example, if you are planning a history file at 3 hour interval then run 12 hour or 24 hour simulation. From the log files (rsl.error and rsl.out) you can figure out the time requires to integrate (first domain 75 second, second domain 75/3; 3rd domain ((75/3)/3); 4th domain (((75/3)/3)/3) seconds.
FYI, the integration times depends on multiple factors, including choice of physical parameterization schemes, grid spacing, time steps, initial and lateral boundary data update frequency. Such as, WRF simulation with Morrison double moment microphysics scheme takes approximately 15~20% more integration time compared to WRF single moment microphysics scheme.
Accordingly, you only gave inner domain grid box numbers and grid spacing but not others three domain sizes. It seems to me your study area is really very big as your inner domain cover an area equivalent to 3,500 square Km.
In summary, your experiment is unique in its kind. So, the only way to get answer of your question is to run a test simulation with your configuration.
I hope the above explanations will be helpful to get your answer.
Thanks
Mostofa Kamal
Hi, I have come across the same issue when trying to move from WRF to WRF-Chem. There is no easy way to estimate the amount of space/memory, as there are just too many variables that can affect it.
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Climate Science
- Climate Change