Differences

This shows you the differences between two versions of the page.

--- wg:ptof:data [2019/06/25 10:46]
brierley [The very very old who knows what ocene]
+++ wg:ptof:data [2019/06/25 12:24]
brierley [Accessing Data at UCL Geography]
@@ Line 88: / Line 88: @@
 ===== Accessing Data at UCL Geography =====
-In preparation for the [[wg:ptof:uclworkshop|May workshop]], some data was collected on a server at UCL Geography.
+In preparation for the [[wg:ptof:uclworkshop|May workshop]], some data was collected on a server at UCL Geography. It contains a curated replica of the ESGF database of the models participating in PMIP. It only has the variables used by the Climate Variability Diagnostics Package (namely monthly tas, ts, pr, psl, mst. CMIP5
-You need to contact [[c.brierley@ucl.ac.uk|Chris Brierley]] to set up an account on this cluster. It contains a
+UCL Geography can make this computer available for server-side compute. You need to contact [[c.brierley@ucl.ac.uk|Chris Brierley]] to set up a personal account on this cluster.
+The directory structure is:
+  * Original monthly data is stored in ''/data/p2f/curated_ESGF_replica/'' and it's sub-directories by GCM and then experiment name. [read-only]
+  * The summary data files (in netcdf) from running the Climate Variability Diagnostics Package are available from ''/data/p2f/cvdp_data/''.
+  * The CMIP6 data donated prior to it being deposited on the ESFG is stored in ''/data/p2f/'' under its model name.
+  * The CVDP summary data files for the CESM Last Millennium Ensemble is also stored in ''/data/p2f/''
+  * All the images and analysis resulting from me having run the Climate Variability Diagnostics Package is stored in ''/data/p2f/cvdp_output''. The most up-to-date of these are avaliable on the web at [[http://www2.geog.ucl.ac.uk/~ucfaccb/PMIPVarData/|the PMIP Variability Database]].
+  * Please use ''~/DATA'' for the results of your own analysis.
+The default version of python is 2.7 on this cluster, because that it required for teaching purposes. Please use ''conda'' to set up your own version. You can access an version of the [[https://pangeo.io/|Pangeo]] stack that includes NCL using the following commands...
+  * ''conda create -f ~ucfaccb/python/environment.yml''
+  * ''conda activate pangeo''
 /* standard page footer */