A Simple Data Structure for Earth Science Model Visualisation NIEeS Visualisation Workshop 29103

1
A Simple Data Structure for Earth Science Model
VisualisationNIEeS Visualisation Workshop29/1/03

• John Laxton

2
Outline of presentation

• What is the Digital Geoscience Spatial Model
  (DGSM)?
• Outline of the Geoscience Spatial Framework (GSF)
  data model
• Outline of the spatial feature/attribute data
  model
• DGSM metadata
• How the GSF will work
• Status

3
DGSM Vision

• The DGSM will encapsulate BGS knowledge of the
  UK and its surrounding continental shelf. It will
  foster a model-centred ethos within BGS, in which
  all geoscience disciplines will contribute to and
  continually enhance the development of a set of
  multidimensional, coherent integrated geoscience
  models.

4
What is the DGSM?

• The DGSM will consist of
• 3D models of the subsurface at a range of
  resolutions
• Models showing different aspects of geoscience
  including
• Lithostratigraphy
• Hydrogeology
• Engineering and geophysical properties
• Geochemical character
• Links to other information including
• Text descriptions
• Images and movies
• Statistical analyses
• Information on the modelling process

5
What is the DGSM for?

• The primary objective of the DGSM is to make
  digital geoscientific information available as
  part of the strategic remit of BGS and for it to
  become the central repository of geoscientific
  knowledge in the UK.
• It has been described as the extension of the
  geological map into 3 dimensions.
• The DGSM will also allow the integration of data
  and aid the development of understanding of the
  geology of the UK.

6
What is the geological map for?

• It represents BGS current best interpretation of
  the geology of an area
• It is publicly available
• It is used as the basis for developing
  added-value products

7
How is geological spatial modelling done in BGS?

• Use a range of commercially produced modelling
  software
• EarthVision
• GoCAD
• Vulcan
• Landmark
• Models can only be used with the modelling
  software that generated them
• This conflicts with DGSM aim of making models
  widely available

8
DGSM Models (1)
9
DGSM Models (2)
10
What is the aim of the Geoscience Spatial
Framework?

• Design, implement and trial a format for holding
  digital models that
• is independent of the originating software
• can feed visualisation applications, including
  over the web
• will preserve the models as modelling software
  and proprietary formats are upgraded

11
Geoscience Large Object Store (GLOS)

• The GLOS holds models in their original software
  formats
• Full richness of original model in GLOS, but at
  present can only be viewed on a workstation with
  originating software
• GLOS dependent on originating software version

12
The GSF and GLOS - How do they relate?

• Loss of model richness and visualisation
  capabilities with GSF
• GSF main vehicle for external dissemination via
  web the new geological map
• GLOS used for internal model exchange,
  integration and development
• GSF and GLOS are complementary

13
GSF Design (1)

• Identified the output from current modelling
  projects the input to the GSF
• Identified the geological sequences to be handled
• Base design on modelling package export formats
  points on a surface
• Describe rock bodies above/below surfaces
• Software to regenerate connectivity

14
GSF Design (2)

• Link to BGS corporate feature/attribute model for
  full object attribution
• Link to metadata for track-back to data and to
  model inference/reasoning
• Link to surface identifier, which will also
  link to GLOS

15
Geological sequences
16
GSF Data Model (1)

• GSF data model holds surfaces (and volumes) as
  points
• Each surface point has up to three aspects
  (above, below, at)
• Each aspect at each point can link to one or more
  features (eg a surface could be a unit boundary,
  and a fault, and a mineral vein)
• The features and their attributes are defined by
  the spatial feature/attribute model
• GSF data model is spatially simple but attribute
  rich
• GSF data model closely linked to corporate data
  model and extensible

17
GSF Data Model (2)

• One XYZ record for each GSF point
• Each XYZ record links to any number of
  aspect/feature records
• Each GSF aspect/feature record links to a record
  in the spatial feature/attribute tables
• Option to define surfaces explicitly
• Option to link GSF points directly to individual
  attributes

18
(No Transcript)
19
Spatial Feature/Attribute Model (1)

• Geology Extent
• Geology Boundary
• Unconformities (major named)
• Linear Landform
• Areal Landform
• Landform Point Observation
• Structural Composite Feature (includes faults,
  folds etc and links to structural database)
• Structural Point Measurement (links to structural
  database)
• Contours

20
Spatial Feature/Attribute Model (2)

• Non-oriented point data (links to existing
  databases such as FOSSLOCS or BRITROCK)
• Major Mineral Vein
• Boreholes
• Major structural element (larger scale than
  Structural Composite Feature)
• Palaeogeological Terrane
• Modelled Geophysical Property Extent
• Hydrocarbon Field Extent
• Tectonic Plate Extent
• Tectonic Plate Pole of Rotation

21
Spatial Feature/Attribute Model (3)

• Geology Extent
• Attribute fields ID type lithology
  lithofacies assemblage depositional environment
  lithostratigraphy seismic stratigraphy sequence
  stratigraphy biostratigraphy chronostratigraphic
  age start chronostratigraphic age end
  chronometric age chronometric age error range
  method of detection alteration type (including
  contact and regional metamorphism) alteration
  age comments.
• Usage Type covered by Geology Extent are
  Solid Geology, Drift Geology, Artificial Ground
  Extent, Mass Movement, Geophysical Horizon e.g.
  base Permo-Triassic reflector, Limits of Source,
  Limits of Seal, Limits of Reservoir

22
DGSM Metadata

• Metadata about the data and sub-sets defined from
  it
• Discovery metadata about the models
• Inference metadata linking the model to the
  data and processes upon which it was based

23
Data sub-sets

• Data sub-set definition includes
• Date of creation of data sub-set
• Inclusion criteria
• Reasons for inclusion criteria
• Exclusion criteria
• Reasons for exclusion criteria
• Statements for whole data sub-set and individual
  records where latter depart from the general

24
Discovery Metadata for Models (1)

• Includes
• A statement of the overall scientific rationale
  of the model
• A statement of overall model purpose
• 3D spatial referencing system, areal extent and
  bounding coordinates
• The introduction of a hierarchy of models

25
Discovery Metadata for Models (2)

• More rigorous use of existing Quality fields,
  including use of DGSM dictionaries
• Quality suitability (overall quality flag)
• Scale/resolution validity
• Usage/fitness for purpose

26
Inference Metadata (1)

• Refers to metadata linking a data sub-set to a
  model or model component
• Two key components of inference metadata
• Process description
• Process quality

27
Inference Metadata (2)

• Process description
• Interpretation
• Interpretation confidence
• Assumptions
• Assumptions confidence
• Interpolation/modelling process
• Interpolation/modelling process confidence
• Statements for whole data sub-set and individual
  records where latter depart from the general

28
How the GSF will work (1)
DGSM Modelling Software (eg Earthvision)
Export format (xyz points)
GLOS
DGSM Metadata population (Oracle)
29
How the GSF will work (2)
Export format (xyz points)
GSF attribution application (ColdFusion)
GSF loading application (SQL Loader)
GSF points and attributes (Oracle)
30
GSF Input Interface

• Two applications built
• ColdFusion application to allow geologists to
  define features to be linked to GSF points
• SQLLoader application to populate GSF tables
• At present work against complete EarthVision
  surfaces
• 345,000 points loaded to GSF for surfaces from
  three models (S Downs Chalk, Cheshire Basin
  Humber)
• Applications being revised to incorporate
  feature/attribute model

31
How the GSF will work (3)
DGSM Metadata (Oracle)
GSF points and attributes (Oracle)
Spatial retrieval
Attribute retrieval
Metadata retrieval
Model selection
Retrieved set of GSF points
32
How the GSF will work (4)
Retrieved set of GSF points
GSF attributes (Oracle)
Triangulation application
VRML generation
FracViewer (?) Exchange3D (?)
Web
VRML Viewer
Web Viewer (??)
33
Design and develop output applications (1)

• Outline design specification has been drawn up
• Spatial selection interface
• Metadata selection
• Attribute selection
• Individual model selection
• Output/visualisation selection

34
Spatial Selection Interface
35
Visualisation options

• Mesh/skeletal surface
• Rendered surface
• Block model
• Point cloud
• Cross-sections
• Export formats eg VRML and XMML

36
Design and develop output applications (2)

• Report on visualisation technology options
• FracViewer software seems best but requires
  some further development
• Alternatives include VRML and Java
• Small application produced to triangulate GSF
  data
• Triangulated data has been imported to a test
  copy of Fracviewer and successfully visualised

37
FracViewer visualisation
38
Four surface GSF output via VRML
39
Status (1)

• GSF data model designed and implemented
• DGSM Feature/attribute logical model agreed as
  core of BGS feature/attribute model
• Feature/attribute model being implemented
  completion end March
• Version 1 of GSF load applications used to
  populate GSF with trial data
• Version 2 of GSF load applications, with link to
  feature/attribute tables, will be built early
  2003/04

40
Status (2)

• GSF retrieval application being developed
• GSF triangulation program implemented
• GSF gridding program implemented
• FracViewer investigated and discussions with
  Fractal Graphics and other vendors on development
  of a web viewer
• VRML output from GSF triangulation

41
Future Work

• Version 2 of GSF load applications to handle
  feature/attributes
• Handling input from full 3D modelling packages eg
  3D Earthvision
• Handling of model delimiters (such as holes)
• Handling of complex surfaces
• Population of attributes that vary over a surface
  (eg younging)
• Development of alternatives to VRML