London Kick Off meeting

From CEO-LD

Report

Scoping points summarised on day-1

Think in terms of a contract between the data provider and the data user

We are concerned with the sharing of data products, not the production chain from satellite to earth station and the curation steps. Although we should be mindful of the limits and opportunities of that process

Structure

We identified three things that the report should include:

  1. Purpose { our robust statement of scope goes here }
  2. Functional Principles - the principles that we're adopting for the product that we're creating
    • Putting data in the hands of people who want to use and manipulate the data without having to be a geo expert
    • Whilst also supporting more expert users/researchers in finding the detailed data and provenance info that they need
    • Be open, we're contributing something distinctive to the W3C/OGC WG
    • No wheel reinvention. Something should be deliverable/actionable, not an academic exercise
    • minimum cost
    • Whatever is delivered should be discoverable, both the work we do and the data it describes
    • [Jeremy / Payam] Needs to be applicable both to real-time data publication and access to data archives
  3. Enabling Access and Delivery

e.g.

  • Identifiers: dataset & distribution … for individual “cells"? for slices & subsets? [VoID]
  • How to discover the data set
  • How to encode the data values - based on the topology of the data? (for exchange, not necessarily storage)
… convert the data to RDF at the point of query?
… binary formats; JPEG, GeoTIFF, HDF5, netCDF etc. - or ascii: RDF serialisation, CSV [on the web], plain-old-JSON [arrays] even! … WebGL
… what formats are directly supported in the browser [user agent - not necessarily a browser]? … CANVAS / Javascript … SVG
… the metadata _must_ be available in browser parseable forms
  • What metadata is needed (for a dataset [GeoDCAT-AP], for a coverage [19123][RDF Data Cube], for a time series [TimeseriesML])- and how to encode (RDF?) / conneg… and [how] can this be indexed by search engines? how to make this metadata both human & machine readable? … what granularity is a dataset anyway?
  • What API should be provided [by a service end-point] for accessing a dataset (or aggregation of datasets)
... geoposition
... time
... observed property (range-type)
... index-space?
... is the API different for Application Developer (non-expert) and Data Scientist (expert who wants flexibility to do it “my way”) ? … are there pathways to move from the non-expert to the expert API?
… how to run Queries against a subset - rather than the whole dataset
  • How to annotate the dataset- or simply make reference to subsets of the dataset … [Opensearch geo & temporal extensions][CHARMe] … “the active volcano is seen in pixels (a,b)-(c,d) in image x”
  • How do I describe the provenance (i) for a processing chain [PROV-O], (ii) for the observation event (O&M, SSN)
  • How do I assert quality [DQV … see DWBP deliverable]

Possible Tests

Are there tests we'd like to see carried out? Subset of large data set as RDF? Would need a use case that just needed a subset

Issues

Metadata about Coverages (specialisation of dataset metadata)

  • ISO 19115/ISO 19123
  • CSW
  • DCAT (GeoDCAT-AP, Dublin Core)
  • VoID
  • Dataset Descriptions: HCLS Community Profile
  • schema.org

Access

  • Access to individual observations
  • Identifiers (URIs)
  • Linkability (to set, slices, down to individual observations)
  • Granularity options?
  • Actually access or just refer to it?

Representation options

  • RDF Data Cube (SDMX, hypercube, slices)
    • Order? Slices as TIFFs?
  • CSV on the Web
  • GeoTIFF??
  • Others?

Scalability

CRS

CEO-LD should provide a view on Issue-28 - whether SDW WG should define a default CRS

Probably define the 'default CRS' explicitly in the vocabulary term

Motivation

Why is access to EO data important, what are the benefits, how can we increase access to that data

Dates

Tuesday 29th (all day) Wednesday 30th (morning)

Participants are welcome to join us for dinner in the evening of Monday 28th as well. Please meet in the reception area of the Covent Garden Travelodge at 18:00.

Meeting Aims

The aim of the meeting is to firstly meet each other, then begin to identify the specific requirements that the CEO-LD project wishes to see met by the W3C/OGC Working Group. This will entail reviewing the WG's Use Cases and Requirements document and, very likely, preparing one or more use cases specifically related to satellite data and the kind of problems that a Linked Data approach can help to solve.

Participants

  • Geoffrey Boulton, CODATA and Edinburgh University
  • Simon Hodson, CODATA
  • Phil Archer, W3C
  • Denise McKenzie, OGC
  • Jeremy Tandy, Met office
  • Simon Agass, Satellite Applications Catapult
  • Maik Riechert, Reading University/MELODIES project
  • Yang Gao, Surrey Space
  • Payam Barnaghi, University of Surrey
  • Jitao Yang, Beijing Language and Culture University/RADI
  • Wang Lizhe, RADI
  • Hu Chunming, Beihang University/W3C
  • Li Jianhui, CNIC
  • Adina Gillespie, SSTL

Not attending

  • Ed Parsons, Google, SDW-WG co-chair
  • Kerry Taylor, Surrey University, SDW-WG co-chair
  • Jon Blower, Reading University/MELODIES project
  • Li Guoqing, RADI
  • Hugh Mortimer, STFC

Draft Agenda

This is a draft and is offered more as a guide of the topics that need to be covered than as a strict timetable.

Tuesday

  • 09:00 Gather, tour de table
  • 09:30 Intro to the project, Phil, Simon & Geoffrey
  • 09:45 Intro to the OGC/W3C Working Group, Phil & Denise
  • 10:00 Review of current/recent work by participants, Simon A, Jitao, others (why is this project interesting to you? What do you hope to get out of it?)
  • 11:00ish coffee
  • 11:30 Context
    • Openess and access
    • Foreseen software
  • Remainder of the day:
    • Review current WG Use Cases, identify gaps
    • Prepare new use cases & requirements

Wednesday

  • Complete notes from Tuesday
  • Steps we can take to increase impact of this work/engage other UK and China stakeholders
  • Agree working methods/expectations/targets
  • Plan for meeting at TPAC, Sapporo 26 - 27 October
  • Planning for Beijing meeting, spring 2016

Venue

The The Royal Society, 6-9 Carlton House Terrace, London SW1Y 5AG. Directions