Hi! Just curious if anyone knows where to find the pressure level for any given CMIP6 dataset in the Pangeo catalog? I see that the ECMWF Copernicus form for accessing CMIP6 data has this option available, but I’m having trouble finding any info on this in the datasets themselves and on the web. I would assume they are all just “single levels” but I would just like to confirm that, and maybe find a place that has that type of meta info for these data as well.
This project is unbelievable - thanks for all you’ve done so far, I hope to be able to contribute more than just questions in the near future!
1 Like
Hello @kmredilla
I have not used pressure level data, but maybe these few info can be of help to navigate the data:
-
to understand what was agreed on from all modelling centers (so not necessary available for each model, or not necessarily yet in the cloud) I strongly recommend you to browse these list of files put together by Naomi The CMIP6 Data for our Hackathon - #2 by naomi-henderson which comes with a table_id for each variable, which essentially specifies you whether they are available at all as pressure level, single level, model level, 3hourly, monthly and so on.
-
the ESGF page is a good place to see what is available from each modelling center - https://esgf-node.llnl.gov/search/cmip6/ - at least to make sure that you are not running after a variable that was not provided by the center itself. I strongly recommend to start broad search and narrow it down slowly. You can use the guide by naomi to figure your way around.
-
This link shows how to search the cloud catalogue. I personally only worked on the google cloud (which should be identical to AWS) and used the csv file approach. you will filter it like a pandas dataframe (there is a couple of example).
Accessing data in the cloud — Pangeo / ESGF Cloud Data Working Group documentation
Start with a broad search and narrow it down. And see what is actually available on the cloud. It won’t be all that is available from the modelling center.
-
when i explored the data I do remember seeing some 6hrPlev data - which are 6hourly values on Pressure level. you can look at one file - again instructions on how to connect to one of the zarr files are on that pangeo website above - and explore the metadata and see what pressure levels are available. In my experience, the data with Plev vertical coordinates have few pressure levels for the vertical coordinate (but of course whether they are sufficient it depends on your needs). Then of course there are some variables that come as one level variables, such as surface temp, air temp, etc.
-
I don’t know the ECMWF copernicus dataset at all. It could (i have no idea) that they interpolated the data from model level to pressure level, and so they have some other type of availability, but I have no idea.
Hope this helps somehow, best!
1 Like
Thanks @chiaral for the thorough reply! That was very helpful. I am seeing that pressure level is usually (always?) a dimension / coordinate variable instead of a data variable, and I think the reason I am not finding the info I was wondering about – official text along the lines of “all data in the Pangeo Google CMIP6 holdings that does not specify a pressure level, or contain a corresponding coordinate variable, can be assumed to be at a ‘single level’” or something – is because it is implied. I’ve been playing around more and checked out one of the air temperature stores and see that it contains a coordinate variable for pressure level, whereas near-surface air temperatures will not. This corresponds with the ECMWF form’s behavior, which will not allow you to select anything other than “single levels” if you have certain variables selected, such as near-surface temperature, precipitation, etc. Thank you again!
I think some of the details you are interested about are on the guide by Naomi. the extended name of the variables will help you decipher things. Also, if you print the attributes after “loading” the stores you will read more details. similarly to a ncdump -h
Glad it helped!