Open Forum Discussion

Hi,

I teach both undergrad and graduate courses that involve climate modeling, and I would like to find software, preferably online, that students could use to actually run a low-resolution GCM and do simple experiments.  Does anyone know of one?  I use MAGICC, which is great as an energy-balance model, but I also want one with atmospheric dynamics and a global spatial regime.  If it has prescribed SSTs or a mixed-layer ocean, that would be OK.  I wanted to use Ed GCM, but it is defunct.  I know of places where you can look up the results from canned simulations, but I want them to change parameters and forcing and see what happens. 

I would think that COMET at UCAR should provide such a learning opportunity to the community.  As far as I can tell they have an Introduction to Climate Models module that describes climate models, but not one set to run on NCAR computers for students.  I know CESM is available, but I want one for a homework assignment that would not involve running the entire complex model.

Any suggestions would be welcome.

Alan Robock, Distinguished Professor
Department of Environmental Sciences         Phone: +1-848-932-5751
Rutgers University                            E-mail: robock@envsci.rutgers.edu
14 College Farm Road            people.envsci.rutgers.edu/robock
New Brunswick, NJ 08901-8551     ☮ twitter.com/AlanRobock

------------------------------
Alan Robock, Distinguished Professor
Department of Environmental Sciences Phone: +1-848-932-5751
Rutgers University E-mail: robock@envsci.rutgers.edu
14 College Farm Road http://people.envsci.rutgers.edu/robock
New Brunswick, NJ 08901-8551 ☮ https://twitter.com/AlanRobock
------------------------------

Hi Alan, 

I have used the University of Hamburg's Planet Simulator (PlaSim) in my graduate courses before. It's a model of intermediate complexity which is easy to compile, has a wide range of resolutions, and can run on the order of 100 year simulations per day on a standard PC. It runs with a GUI that gives you the option to visualize the output in real-time, which is fun for teaching students modeling and/or demonstrations, but it also has the capability to save the output for analysis. The GUI also allows students to quickly change the basic parameters and it's fairly easy to show how to alter more complicated options in the namelists. In my climate theory course I have them use it to run climate sensitivity experiments and reproduce some of the basic results from early GCM papers (e.g., Manabe and Weatherald 1975, Hansen et al. 1984). Here's a link: https://www.mi.uni-hamburg.de/en/arbeitsgruppen/theoretische-meteorologie/modelle/plasim.html My only complaint about PlaSim is that the standard output is in binary and requires a postprocessor to do any analysis which can be a little challenging to set up.

For undergraduate classes, especially lower level ones where they have little to no command line experience, it's more challenging to think of a good tool. I have not used it but there is something called EzGCM (http://ezgcm.org/ezgcmdev/#/login) which was described to me as EdGCM but run in the cloud. There is a cost associated with it, but I haven't explored far enough to see what it is. 

I love the idea of more accessible, educational GCMs that could possibly be run on remote machines and perhaps connected with COMET. CESM can be run in a docker environment that makes porting it easier, but there still is complexity in getting the docker environment right that would be a barrier to students at certain levels and it would still be too complex for some course levels. It would be amazing to have an open source replacement for EdGCM. 

I'd be curious to see if anyone else has ideas about this, this kind of tool would be extremely useful for teaching climate modeling. Thanks for bringing this topic up! 

Be glorious,
Ross

------------------------------
Ross Dixon
University of Nebraska-Lincoln
------------------------------

A very simple question about a very, very simple climate model: has the Manabe and Wetherald '67 radiative-convective model ever been made into something that can be used educationally, where the very few parameters in it could be tweaked, etc., to get different results?  It should be around somewhere, but I have not yet located it.  Don't want to reinvent the wheel.  Thanks, John Knox

------------------------------
John Knox
University of Georgia
Athens GA
------------------------------

Original Message:
Sent: 11-29-2023 09:43
From: Ross Dixon
Subject: Can you recommend a GCM climate model to use as a class exercise?

Hi Alan, 

I have used the University of Hamburg's Planet Simulator (PlaSim) in my graduate courses before. It's a model of intermediate complexity which is easy to compile, has a wide range of resolutions, and can run on the order of 100 year simulations per day on a standard PC. It runs with a GUI that gives you the option to visualize the output in real-time, which is fun for teaching students modeling and/or demonstrations, but it also has the capability to save the output for analysis. The GUI also allows students to quickly change the basic parameters and it's fairly easy to show how to alter more complicated options in the namelists. In my climate theory course I have them use it to run climate sensitivity experiments and reproduce some of the basic results from early GCM papers (e.g., Manabe and Weatherald 1975, Hansen et al. 1984). Here's a link: https://www.mi.uni-hamburg.de/en/arbeitsgruppen/theoretische-meteorologie/modelle/plasim.html My only complaint about PlaSim is that the standard output is in binary and requires a postprocessor to do any analysis which can be a little challenging to set up.

For undergraduate classes, especially lower level ones where they have little to no command line experience, it's more challenging to think of a good tool. I have not used it but there is something called EzGCM (http://ezgcm.org/ezgcmdev/#/login) which was described to me as EdGCM but run in the cloud. There is a cost associated with it, but I haven't explored far enough to see what it is. 

I love the idea of more accessible, educational GCMs that could possibly be run on remote machines and perhaps connected with COMET. CESM can be run in a docker environment that makes porting it easier, but there still is complexity in getting the docker environment right that would be a barrier to students at certain levels and it would still be too complex for some course levels. It would be amazing to have an open source replacement for EdGCM. 

I'd be curious to see if anyone else has ideas about this, this kind of tool would be extremely useful for teaching climate modeling. Thanks for bringing this topic up! 

Be glorious,
Ross

------------------------------
Ross Dixon
University of Nebraska-Lincoln
------------------------------