RD2 20231010 meeting

Participants

Alena Trojakova Benedikt Strajnar Carlos Geijo Florian Meier Gunther Haase Jana Sanchez Magnus Lindskog Martin Imrisek Martin Ridal Martina Tudor Maud Martet Mats Dahlbom MiNe Patrick Moll Phillip Scheffknecht Siebren De Haan Suzana Panezic

DA2.1 Local implementation of radar DA

Gunther’s presentation on OPERA 5 and 6

OPERA 5

Nimbus is the new OPERA production line to replace Odyssey production for rain rate composite and single-site radar data for NWP. The developments are made by KNMI and geosphere. Wind profiles production was planned in OPERA 5 bu ithas been postponed to opera 6. Currently, the pre-operational phase is ongoing at geosphere but the operational suite will not be ready for end of 2023.

+Evaluation of composites:+ Timing is ok, completeness of the composite is quite ok and quality of composite will be studied in the next moth (maybe beginning of 2024). A longer period of data is needed.

+Quality control volumes:+ Data are available on an ftp server (equivalent to OIFS). If you want to have access to this server, you can send an email to Lukas Tuchler at Geosphere. Currently, 5min and 15 min volumes available for testing but there are problems to gather wind and reflectivity volumes. An OPERA meeting between OPERA, users and developers is planed (11/10/2023 9h). Nimbus was supposed to be fully operational at the end of 2023. Unfortunately, it won’t be the case, the testing phase will perhaps begin in November. An extension of Odyssey production for at least 3 months, maybe 6 months is asked but there are budget issues (under discussion at OPERA level).

+ODE server: data for NWP for research+ There are limitations on future due to budget reduction. The servers won't shut down until Nimbus is ready.

OPERA 6

Currently, discussions on budget are ongoing. Development will be focused on removal of non meteorological echoes, improvement of satellite data filtering, dealiasing radial winds, and extension of data exchange to dual polarisation moments.

Discussion

It is required in Nimbus to have similar data than Odyssey (not yet done!). More information will be available after OPERA, users and developers meeting.

Current developments on these subjects

• Slovenia: work on dealiasing for radial winds (validation and impact study), superobbing in HOOF (to be tested), monitor complete volume datasets from European countries. There are drying problems during the assimilation of radar reflectivity but new developments are available in cy49 leading to improvement in precipitation forecast but not on cloud and humidity fields. Also they will perform a monitoring of European radars to be compared with the one ECMWF is doing.
• HIRLAM = problem of spinup with cycle 43 and radar assimilation in MetCoop domain. Too little rain/humidity in the first hours. Radar has been removed from MetCoop pre-operational suite. This issue is not visible at AEMET.
• CHMI = work on use of operational radar data. Work on radial winds is ongoing
• Spain = adapt to the new files, parallel test between ODE and OIFS (in order to have Portuguese data), assimilation of radial winds: from passive to active in operational run.
• Austria = use of nimbus data+bigger domain+ LHN (maybe)

DA2.2 Ground-based GNSS observations

Martin presentation of EUMETNET coordination work (SMHI, MF, KNMI)

A report will be presented to Eumetnet (12/10/2023). Comparisons of experiments assimilating ZTD, ZTD gradients and STD in MetCoop and in AROME-France are performed. ZTD gradients assimilation works technically and the same for STD. STD and gradients data has been created by NGAA and IGN for the time period requested for this study, but they are not produced regularly.

The main problem is to have access to data. Discussion on file format are on-going.

DA2.3 Rain-sensitive observations

Phillip's presentation

Assimilation of microlink data has been presented, using 1d+3dvar method like P. Lopez in IFS and Zahra in Morocco. This method allows to adjust humidity in both directions. The current suite uses obsoul to get data in the model. And although the data delivery ended when the project finished, this method could be used for other types of observations.

Lidar

Currently, nobody works on it.

DA2.5 New technique for assimilating radar data

Maud’s update

Direct assimilation of reflectivity is developed at Météo-France in a 3DEnVar AROME-France assimilation system with hydrometeors in the control vector. Concerning dual-polarisation data, one limitation is the microphysics scheme that is not precise enough to be able to assimilate them properly.

Next meeting

The date of the next meeting will be discussed when working weeks topics will be fixed in order to have a meeting before a working week.

AOB

Should we use more slack or other ways to discuss? As the rest of the groups: Slack for quick communications and issues (remember that messages are removed after 90 days there) and Wiki (link to RD2 page) for more permanent issues and documentation.