These minutes have been copied unedited from the Huddle site
In previous meeting (15th Dec 2009) we agreed a number of points that provide scope to our activities:
- that our objective is to exchange data easily between communities including: operational meteorology, aviation, INSPIRE, climate & atmospheric scientists
- that we should use topology to classify / categorize our domain model ... extending this to enable description multiple members of a single ensemble analysis
- that rights management / ownership and quality management were becoming increasingly important factors to our community of interest
- that we should focus on a 're-usable core' of feature types that can be specialized as necessary for use within domains that stricter controls over data exchange (such as operational meteorology within aviation)
Furthermore, we agreed to use the INSPIRE methodology for developing our conceptual model. This is detailed in INSPIRE document D2.6 (http://inspire.jrc.ec.europa.eu/reports/ImplementingRules/DataSpecifications/D2.6_v3.0.pdf
The first step in the methodology is to develop the use-cases for interoperability.
I propose that the main body of this meetings discussion should focus on identifying the target use-cases that we will use to develop the conceptual model.
From a WMO perspective, I note that there is a compelling need to focus on the development of data models required to enable retirement of 'traditional alphanumeric codes' [TAC] within Aviation OpMet
that are compatible both with existing 'table driven code forms' [TDCF] and ISO TC/211 conceptual model for geospatial information. The target here is to demonstrate to WMO CBS in Dec 2010 that the correct approach is being adopted - we do not need to have completed development of the complete conceptual model! Whilst considering replacement of TAC, we must also look to the future for Aviation OpMet
, suggesting a move away from exchanging data products
to facilitating the extraction of data from 4-dimensional (x, y, z & T) 'data-cubes' for input alongside other aeronautical information into, say, flight routing algorithms.
I also note that INSPIRE has a milestone to complete the development of version 1 of the Annex 3 theme data specification by 2010Q4.
/ NNEW and SESAR are also likely to have milestones in place.
The goal of the meeting will be to establish a candidate set of use-cases to develop the conceptual model against. If insufficient stakeholders are present to agree the list of use-cases then we will attempt to identify stakeholders to review the list offline.
It is important to gather additional proposals from members of the community to ensure that the use-cases are representative of our concerns.
1) Update on actions
2) Update on selection process for INSPIRE Thematic Working Groups
3) Discussion of use-cases to drive development of conceptual model
4) Any other business
5) Arrangements for next meeting
My thoughts and notes on sample Use Cases
Use cases for conceptual modelling
Architectural Framework? Strategic, informal cases preferred.
Institutional vs activity goals/Ucs
Presumption for cross organisational use cases
1. WMO data exchange, for Inspire 'as-is' analysis
, VAAC, etc
Radar European Composite
Satellite, radar, fields, obs plot composite.
2. Calibration of remote sensing
3. Boundary conditions for Local Area Models for a crisis area
4. Aviation - weather cube forecast, with updated forecast issued mid flight based on new forecast
Aviation - new route because of new obs (e.g. volcano en route, or terminal airfield SigWx
Aviation - use of ensembles to improve CAT forecasting.
5. Targetted observation in mid Atlantic from both drop sondes, ASAPs, and ocean submarines and buoys for hurricane forecasting
I have added a new page to the OGC Twiki (http://external.opengis.org/twiki_public/bin/view/MetOceanDWG/MetOceanUseCases
), along with a couple simple institutional use cases and the NextGen
system use case document.
Jeremy Tandy (Metoffice), Andrew Woolf (STFC), Dominic Lowe (STFC), Manuel Fuentes(ECMWF), Chris Little (Metoffice), John Schattel (NOAA), Pierre Kerhevé (WMO), Marie-Francoise Voidrot (Météo-France), Frédéric Guillaud (Météo-France), Stefano Nativi (CNR), Aaron Braeckel (NCAR)
MINUTES: (1) Update on actions
Note: With these minutes, I'm instigating an action (and decisions) list for traceability
[Action-id/Who]: (what) status
[A1/Andrew Woolf] (Subversion repository) Has been set up by OGC staff but with wrong name (metec.dwg); currently changing it and should be available soon.
[A2/Andrew Woolf] (users of geometry/topology as classification axis) Action still open.
No volunteers for future minutes, but offline Dominic Lowe agreed to do this for future (MANY THANKS DOM!).
JT read through the 'PRE-MEETING NOTE', which identified a possible use case around OPMET (ref. the discussion on extracting aviation-relevant data from 4-d data cubes). AB indicated that SESAR/NNEW do indeed have similar use cases available.
MINUTES: (2) Update on selection process for INSPIRE Thematic Working Groups
AW reported that organisations can register as stakeholders: either 'Spatial Data Interest Communities' (SDICs, including non-European) or 'Legally Mandated Organisations' (LMOs) (see http://inspire.jrc.ec.europa.eu/index.cfm/pageid/7
). Such stakeholders can then nominate experts to participate in data specifications development.
The nomination call was extended until 8 January 2010 for Annex III (which includes Meteorology, Oceanography, Atmospheric Conditions).
MINUTES: (3) Discussion of use-cases to drive development of conceptual model
CL: It is important to balance Use Cases for current applications with potential future scenarios (e.g. 4-d weather-cube).
JT: Yes, Use Cases must still be relevant for today's context; e.g. WMO in 5-10 years will more than likely be exchanging data as today.
CL has prepared some use cases around: routine observations and forecasts, routine warnings (e.g. tsunami, volcanic ash, etc.), traditional observations (satellite, weather, data, etc.).
[A3/Chris Little] CL will post details of his use cases.
AW described the INSPIRE hydro use cases. In principle they should have been supplied by the European Commission, but it was left to individual Thematic Working Groups to define their own. For the 'Hydrography' theme (actually hydrology!), these were identified as: Mapping, Modelling and Spatial Analysis, and Reporting. The final model reflected these by having different views of hydrography: a 'physical waters' view (with watercourses, standingwaters, etc. as unrelated geometric objectswith attributes), a 'network' view (with nodes and edges, suitable for hydro flow modelling), and a 'reporting' view with feature types relevant for reporting obligations (e.g. hydro reporting units under the EC Water Framework Directive).
AW described the CSML use case, which was to enable harmonised data exchange for the British Atmospheric and Oceanographic Data Centres, which have completely independent (and different) data management infrastructures. The development unintentionally used an approach similar to INSPIRE by also asking stakeholders for what they regarded as important 'data classes' from a conceptual perspective (similar to the INSPIRE 'as-is' analysis).
The Use Cases of CL, and the data we exchange today, outline minimum requirements.
While there is an overlap between some use cases and today's exchange, it is also useful to keep these separate -c.f. the INSPIRE difference between use cases and 'as-is' analysis.
CL suggested we could start with high-level scenarios, and drill down to more detailed use cases.
AB agreed we should start at a top-down strategic level, but high-level scenarios aren't necessarily enough. The DoD
Architecture Framework standard may be useful for formal capturing of use cases; also for capturing multi-level use cases.
JT: Use cases are a tool to get us started, rather than an end in themselves, prefer to remain informal (e.g. textual descriptions capturing scenarios). UML doesn;t add a lot for UCs because we are not doing model-driven design. Agile development has the concept of 'user stories' (a paragraph/words that describes the 'intent' of user interaction).
MINUTES: (3 ctd.) Discussion of use-cases to drive development of conceptual model
MF: Examples of other use cases could be useful.
[A4/Aaron Braeckel] Will send details of Nextgen/NNEW use cases
Use case examples should be added to the Metoc wiki.
Are there any specific requirements for, or outcomes from, the data model that we can identify already?
MF: Agreed with CL earlier (routine obs/forecasts etc.), also exchange with space agencies (e.g. for cal/val), routine weather obs, use of NWPs as boundary conditions for other models.
PK: Nothing specific in mind, WMO resolution 40 provides constraints on exchange
AB: Need to think first of strategic top-level organisational requirements. It's not 'getting satellite data around', more 'goals of stakeholders to meet their business requirements now and in the future'.
Some consolidation will be needed on candidate use cases to fit in an overall structure. This can be done sequentially: first collect institution-level use cases, then drill down in a second round.
AB will send use cases as per above action, furst thinking about types of stakeholders (e.g. research suppliers, nextgen users, etc.); aiming to think about institutional goals for data exchange.
JT: An insttiutional goal of 'routine observation exchange' is to provide input for daily analyses and verification/QA.
CL: Crisis-area modelling also identifies non-traditional data flows.
MF: Institutional goal is to provide the best information possible for security of people and infrastructure; so need to be able to adapt system for crisis situations.
JT: Does a crisis require large-scale gridded information, or simplified warnings?
MF: Gridded data is used to identify a crisis situation, but the warning itself is a simple message or graphic.
JT: Will also depend on the particular consumer of the warning information.
MINUTES: (3 ctd.) Discussion of use-cases to drive development of conceptual model
DL: When talking about a 'common' data model we should be careful of going too far down individual institutional requirements, also need to look at the cross-organisational issues.
SN: One can accommodate different systems/requirements in a 'system of systems' approach, i.e. keep local autonomy starting from the present situation, but incorporate a degree of flexibility. Another aspect of INSPIRE is the cost-benefit analysis (i.e. build on existing systems, minimising costs, bugt maximising interoperability benefits).
MF: Could such criteria be used to filter relevant use cases?
SN: Not really, it rather indicates the importance to consider interactions at the system/enterprise level.
JT: It is a good point to bear in mind our starting point. E.g. Gil Ross mentioned that the INSPIRE data specification for Coordinate Reference Systems references netCDF and GRIB - there may be scope to continue to use existing technologies but with an 'interoperability wrapper'. This type of requirement is important. What about GEOSS?
SN: GEOSS AIP may have some helpful use cases (e.g. air quality), also GMES has some nice use cases.
JT: GMES/GEOSS etc. and other use cases should be added to the OGC twiki under today's meeting links.
CL: Already uploaded his notes, will need expanding out for the twiki.
JT: First need to collect/collate, then to review/rationalise teh use cases.
[A5/Jeremy Tandy & Andrew Woolf] Create top-level link for collecting use cases for development over next two weeks.
[A6/all] Add use cases.
MINUTES: (4) Any other business
AB: The UCs should go as top-level link on wiki?
MFV: There is already a page for conceptual modelling, can be changed to a top-level page.
[A6/Jeremy Tandy] Will make a post on huddle once the change to wiki structure is made
MINUTES: (5) Next meeting
Already set up for 26 January, 15:00 UTC.
MINUTES: (6) open actions
[A1/Andew Woolf] Subversion repository
[A2/Andrew Woolf] Users of geometry/topology as classification axis
[A3/Chris Little] Post details of his use cases
[A4/Aaron Braeckel] Post details of Nextgen/NNEW use cases
[A5/Jeremy Tandy & Andrew Woolf] Top-level link on wiki for adding use cases
[A6/all] Add use cases
[A7/Jeremy Tandy] Post on huddle once wiki structure is changed
[NEXT FREE ACTION NUMBER: A8]
- 09 Feb 2010