I am not familier with NWP and I have recently found the MetEd website.I want to know what computer resource is needed to run numerical models? can i do this on pc?

here are links to some of the nwp models - most of them are open source.

http://www.mmm.ucar.edu/mm5/workshop/

http://www.caps.ou.edu/ARPS/index_flash.html

http://vortex.atgteam.com/

http://www-paoc.mit.edu/cmi/default.htm

http://www.mmm.ucar.edu/wrf/users/do...t_source2.html

You could try - however you would need to make some serious cut backs in gridspacing and variables carried (you probably won't need data assimilation - this is a serious saving in resources). Some meteorological services provide a portable version of their code which can be configured for use with a workstation - although these codes are free they are copyrighted!

For a typical NWP simulation (performed operationally) there are of the order 4 million grid points for which you need to store 3 velocity components, pressure, temperature, humidity, moisture, solar radiation (long and short wave) various microphysics parameters (cloud water, ice, etc), ozone and some other chemical species, etc ---> One very big and fast computer!

Most of the effects listed above can be switched off (which you would need to do to run it on a PC). You'll also need forcing fields (solar radiation, orography datasets, land vegation etc) before you can run the model - you'll also need to build a startdump which is consistent with your model setup to initialize the simulation.

The data output from the model is likely to be in a nonstandard form (depending on where you get the code from) and so you'll need some software for extracting the data from the diagnostic files for plotting purposes.

Don't let my above rambling put you off though - if you have the time and are interested in these things it's definitely a worth while exercise!

Tom.

P.S. if you're new to meteorology it's worth looking at the book "Atmosphere-Ocean dynamics' by Adrian Gill - it'll also help you understand what the model is doing.

[INSYS]

Originally Posted by Tom
;33138

You could try - however you would need to make some serious cut backs in gridspacing and variables carried (you probably won't need data assimilation - this is a serious saving in resources). Some meteorological services provide a portable version of their code which can be configured for use with a workstation - although these codes are free they are copyrighted!

For a typical NWP simulation (performed operationally) there are of the order 4 million grid points for which you need to store 3 velocity components, pressure, temperature, humidity, moisture, solar radiation (long and short wave) various microphysics parameters (cloud water, ice, etc), ozone and some other chemical species, etc ---> One very big and fast computer!

Most of the effects listed above can be switched off (which you would need to do to run it on a PC). You'll also need forcing fields (solar radiation, orography datasets, land vegation etc) before you can run the model - you'll also need to build a startdump which is consistent with your model setup to initialize the simulation.

The data output from the model is likely to be in a nonstandard form (depending on where you get the code from) and so you'll need some software for extracting the data from the diagnostic files for plotting purposes.

Don't let my above rambling put you off though - if you have the time and are interested in these things it's definitely a worth while exercise!

Tom.

P.S. if you're new to meteorology it's worth looking at the book "Atmosphere-Ocean dynamics' by Adrian Gill - it'll also help you understand what the model is doing.

Hi
I am wondering if it is possible to do numerical weather predictions for regional and global models in ANSYS Fluent or CFX.