DE33052 WP5.2

Tasks and updates from weekly meetings

daily production with 3D fields, + cw, ci, cloud fraction, 9 km, 137 levels, on ATOS /scratch/, nfbcli is recommended to be 2 (Maria, Neva)

20221122: The daily production has been stable since Nov 15 (Neva); data is syncronised over to LUMI (Ulf)

Maria: what is the plan with DTE about offline surfex, or direct coupling, or coupling via operational LAM?
20221122: Initialisation of surfex fields from IFS: ALADIN/ALARO: via FULLPOS; Harmonie-arome via PREP. Ulf suggests to take the FULLPOS route with DEODE

The PGD ML group in WP3.2 aim is to improve the ECOCLIMAP-SG (ECO-SG) cover database and increase the resolution from 300m to 60m.
We need to split the global PGD maps to both speed up the physiography setup for the hectometric domains and to be able to move to higher res. input data.
Status (20221129):
- LAI and Albedo tiled up (10N-90N/90W-90E) at 150m res.
- ECO-SG Cover map tiled up (10N-90N/90W-90E) at 60m res.
- Bolli has made updated ECO-SG maps for Greenland (90m), Iceland (25m), Faroe Islands (90m) and is working on Svalbard (30m).

Procedure to do climate generation
Harmonie-arome procedure in the hirlam wiki page.
- AROME/ALARO CSC currently does it on MF HPC?
20221122:
- database inventar: collection of database files/scripts used for climate generation on MF server (Neva, Maria D) and those used in Harmonie-arome to achieve a complete dataset, with aim for enabling of climate generation on on ATOS/LUMI
- common climate generation scripts shall be targeted in synergy with WP5.1 (Kasper, ?)

Discussion about data format selected for crowdsourced data (netatmo, smartphone) from the ACCORD DA working week

20221122: Aim for a discussion on 20221129 With Bene participating
20221129: Bene reports preliminary setup using OPLACE data as baseline to display data availability. Suggestion is to try input of diverse data source to get a maximum picture and to examine about ownership issues. As next step, real time data from ECMWF (ATOS: /ec/vol/msbackup ) will be tested. Future step include a merge facility to check redundancy. For data types involving VARBC the strategy is to collect them first before worrying about VARBC.