Open adoerenbecher opened 1 month ago
Sorry, I've forgotten to include a diagram that shows how the TSEN system is built - especially the vertical distance between the sensors. Here is the picture. !
https://github.com/wmo-im/tt-tdcf/wiki/2024.10.16.tt.tdcf notes:
Following the meeting on Oct. 16, I formalised the sequence of this proposed template. The description above is more or less a list of content. Colleagues explained to me how e well-defined sequence limits the number of new entries/descriptors to be created.
For example, it is not necessary to define a height in relation to the station for each parameter/sensor; it is sufficient to use a generic descriptor and to use the sequence "pressure", "temperature", "humidity" and "winds" to specify the vertical positioning of each sensor under the balloon.
The same logic applies to the sampling description. In the case of Stratéole 2, some raw data are acquired at high frequency and are averaged subsequently. The descriptors 0-25-170 and 0-25-171 are used to give this information. It is updated for each sensor, as each sensor may have its own acquisition protocol.
As a consequence, the measurement section should better read as follows.
CATEGORY | ELEMENT NAME | TABLE REFRENCE | UNIT | SCALE | REF. VALUE | DATA WIDTH | STATUS | COMMENTS | EXAMPLE |
---|---|---|---|---|---|---|---|---|---|
Time significance | 0-08-021 | Code Table | 0 | 0 | 6 | existing | 2 | ||
Height of sensor wrt station | 0-07-077 (new) | m | 0 | -400 | 12 | to be proposed | 0 | ||
Sampling interval | 0-25-170 | s | 0 | 0 | 10 | existing | 2 | ||
Sample averaging period | 0-25-171 | s | 0 | 0 | 10 | existing | 600 | ||
Pressure | 0-07-004 | Pa | -1 | 0 | 14 | existing | |||
Geopotential height | 0-10-009 | gpm | 0 | -1000 | 17 | existing | |||
Height of sensor wrt station | 0-07-077 (new) | m | 0 | -400 | 12 | to be proposed | -10 | ||
Sampling interval | 0-25-170 | s | 0 | 0 | 10 | existing | 2 | ||
Sample averaging period | 0-25-171 | s | 0 | 0 | 10 | existing | 600 | ||
Temperature/air temperature | 0-12-001 | K | 1 | 0 | 12 | existing | |||
Standard deviation of temperature | 0-12-051 | K | 0 | 0 | 10 | existing | |||
Height of sensor wrt station | 0-07-077 (new) | m | 0 | -400 | 12 | to be proposed | missing | ||
Sampling interval | 0-25-170 | s | 0 | 0 | 10 | existing | missing | ||
Sample averaging period | 0-25-171 | s | 0 | 0 | 10 | existing | missing | ||
Water vapour mixing ratio | 0-13-002 | kg/kg | 5 | 0 | 14 | existing | missing | ||
Standard deviation of mixing ratio | 0-13-165 (new) | kg/kg | 5 | 0 | 14 | to be proposed | curr. vacant | missing | |
Height of sensor wrt station | 0-07-077 (new) | m | 0 | -400 | 12 | to be proposed | 10 | ||
Sampling interval | 0-25-170 | s | 0 | 0 | 10 | existing | 60 | ||
Sample averaging period | 0-25-171 | s | 0 | 0 | 10 | existing | 600 | ||
Wind direction | 0-11-001 | degree true | 0 | 0 | 9 | existing | |||
Wind speed | 0-11-002 | m/s | 1 | 0 | 12 | existing | |||
Standard deviation of wind direction | 0-11-049 | degree true | 0 | 0 | 9 | existing | |||
Standard deviation of wind speed | 0-11-050 | m/s | 1 | 0 | 12 | existing |
The template is open to the introduction of other parameters in relation with e.g. atmospheric chemistry.
I attach here a PDF document to better show the clarified sequence for this template.
As noted by a colleague within DMI, the data width for the proposed new 0-01-158 ("Type of balloon envelope") descriptor appears to be incorrect. It is listed as 16, when a look at the corresponding proposed code table indicates the data width should be 4 instead.
https://github.com/wmo-im/tt-tdcf/wiki/2024.11.12.tt.tdcf notes:
Initial request
Rationale: Field campaigns with long duration drifting balloons are quite common/frequent. Sometimes high-value data are available in real-time and could be used in numerical weather prediction applications (e.g. data assimilation).
Drifting balloons are a world unto themselves and a very wide variety of measurements can be collected using them (both in-situ and remotely sensed data, e.g. radio-occultation) because these are primarily vectors. These vectors are chosen for their ability to transport sensors to atmospheric regions that would otherwise be difficult to access.
These drifting systems are very often fitted with sensors to measure the close environment of the balloon, and these are the data of interest here.
With this issue we propose to set up a new BUFR template to allow the dissemination of in-situ data (at the flight level) to the meteorological community. Such a BUFR template would be a trade-off between aircraft or radiosonde messages, involving the definition of new tables as a few novel concepts should be defined.
Purpose: The requestor (Météo-France Research Department, aka CNRM) has some experience in using observations collected by balloons operated par CNES (French National Space Agency) in various field campaigns such as CONCORDIASI, Stratéole-2 (stratospheric data in the global model) or HyMeX (marine boundary layer above the Mediterranean in the regional model). During these campaigns the data were made available using FTP repositories. Having an official BUFR template dedicated to these balloons data would also help in the dissemination of the data to the community.
The Stratéole-2 is not over and a last (third) campaign will take place at the end of 2026 with flights circumnavigating the tropical tropopause layer (TTL) for months with a flotilla of about 20 balloons drifting at either 18 or 20km of altitude. Many experimental systems will collect research data, but the TSEN system (IPSL/LMD, France) is able to collect high quality data and send them in real-time through satellite link. Météo-France participates in the Stratéole-2 effort and is ready to disseminate the data on the global telecommunication system if WMO allows it, and having a proper BUFR template is part of it.
At CNRM, we have designed a preliminary local version of a BUFR template to make tests with the data of the two first campaigns of the Stratéole-2 project. This dedicated BUFR format allow both the IT system and data assimilation systems to handle balloon data properly. For these data which are neither radiosonde, neither aircraft data, it is crucial to properly control black or whitelisting, observation uncertainties or thinning procedures.
Implementation plan: Here we simply submit our attempts to build a sensible template (3 11 012) to your expertise. We have used new code for descriptors corresponding to the quantities that were not already defined in the BUFR documentation. We hope that these do not clash with recent descriptors. Depending on the decisions of the WMO experts, we can adjust the preliminary template we have build and provide WMO with a sample of real data collected during the past Stratéole-2 data. As soon as the BUFR template is validated we will ask WMO for an authorization to disseminate the Stratéole-2 data. If authorized, we will prepare the data flow in order to be ready in September 2026.
Amendment details
The tables below describe the descriptors we are proposing for the new BUFR message model for data from drifting balloons. We have tried to follow the format used in the official BUFR documentation. The last column gives examples of values for data from the Stratéole-2 campaign.
Most of the descriptors used here already exist.
However, new descriptors need to be created, in particular for metadata describing the balloons and also how the data is collected. In the case of Stratéole-2, the measurement station is not compact like a radiosonde and the measurement points are distributed vertically: it is therefore necessary to be able to describe this staggering in relation to the point where the altitude is measured (GNSS). We have put a ‘new’ label next to the descriptors to be created.
We include the necessary structure for humidity information, even though Stratéole-2/TSEN data do not have humidity. But in case it may happen with other balloons (e.g. encoding for Windborne data) the structure is already there.
The following section (publication) lists the descriptors to be created.
Here follow the descriptors for the data :
As we introduce new Code tables... Here are the information for the code figures to be used.
Comments
We have a PDF document about this BUFR template proposal. It is available on demand.
Requestor(s)
DOERENBECHER Alexis, Météo-France, Reaserche Center (CNRM) and Stratéole-2 project (alex.doerenbecher@meteo.fr) POURRET Vivien, Météo-France, Reaserche Center (CNRM) and Stratéole-2 project (vivien.pourret@meteo.fr)
Stakeholder(s)
The amendment has been designed to not change (hopefully) the existing BUFR matter, but only to add new features to the BUFR system. All the research and operational numerical weather prediction centers might be impacted as soon as they try to decode balloon data that would have been disseminated by the Météo-France/Stratéole-2 project on the global telecommunication system by the end of 2026 (see "initial request" section for more details on the implementation plan).
Publication(s)
Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Table D, Class/Category 11, add BUFR descriptor 3 11 013 (template) Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 01, code table 0 02 011: add TSEN station Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 01, add BUFR descriptor 0 01 156 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 01, add BUFR descriptor 0 01 157 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 01, add BUFR descriptor 0 01 158 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 01, add BUFR descriptor 0 01 159 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 02, add BUFR descriptor 0 02 192 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 02, add BUFR descriptor 0 02 193 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 07, add BUFR descriptor 0 07 077 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 07, add BUFR descriptor 0 07 078 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 07, add BUFR descriptor 0 07 079 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 07, add BUFR descriptor 0 07 080 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 08, add BUFR descriptor 0 08 100 Manual on Codes (WMO-No. 306), Volume I.2, BUFR, Class 13, add BUFR descriptor 0 13 165
Expected impact of change
MEDIUM
Collaborators
Collaborators are the principal investigators (PI) in the Stratéole-2 project. Stratéole-2 project is a French-American project to better understand atmospheric mechanisms at stake in the Earth climate system, focusing on the Tropical Tropopause Layer and hydration processes in relation with deep convection. Stratéole-2 has 2 field campaigns involving instrumented stratospheric balloons operated by CNES (plus a test campaign). Scientific payloads may vary from one balloon to another and most of these are cutting edge systems to collect unprecedented dataset. However, all balloons are fitted with high quality instrumentation to measure pressure and temperature at flight level and also estimate horizontal wind (TSEN system, IPSL/LMD). The data collect by the latter are available at high-frequency and in near-real-time. These data are of high value global numerical weather prediction activities. All other data are available on open database after the campaigns. The PI for the TSEN system is Albert Hertzog (IPSL/LMD). He is responsible for the raw data collection and preprocessing. The CNES chief manager of the balloon operations is Stephanie Venel.
References
Here follow a few references about Stratéole-2. TSEN website: https://www.lmd.ipsl.fr/en/instruments-lmd/tsen-2/ Stratéole-2 in EOS: https://eos.org/science-updates/around-the-world-in-84-days Stratéole-2 white paper: https://www.lmd.polytechnique.fr/VORCORE/Documents/US%20Involvement%20in%20the%20Strateole%202%20Balloon%20Campaign.pdf
Validation
A prototype of this BUFR template for balloons has been built at Météo-France (research center) on the basis of local tables that can be used with EcCodes.
Data assimilation experiments in the ARPEGE global model have thus been carried out and the impact of TSEN data from the 2 Stratéole-2 campaigns that have already taken place (test campaign and campaign n°1) has been measured. A scientific article on the results is being prepared and is not yet ready for distribution.
The results were presented at the Stratéole-2 conference in Boulder (USA) in June 2023. The presentation is available on request.