The GEON IDV

1

• The GEON IDV
• ( GEON Integrated Data Viewer )
• for 3D/4D Visualizations of Data from Distributed
  Sources
• Charles Meertens and Stuart Wier
• UNAVCO/GEON

2
Outline

• Data access and distribution
• IDV Integrative Data Viewer (3D/4D Visualization,
  WMS, OPeNDAP)
• Integration
• True 3D Demo
• Tutorial install and learn the GEON IDV

3
GEON 4D Data Access and Visualization

• An ongoing GEON effort is to address 4D (xyzt)
  representation of earth science datasets and
  models in a grid computing environment. Current
  desires and approaches include
• Capable volume-time Integrative Visualization
  tools Enhancing the Unidata IDV Java Application
  for earth science data
• 4D (and multi-parameter) Data Model Adopting
  netCDF (used by IDV and soon by ESRI and GMT).
  Common Data Model effort at Unidata
• Data Discovery GeonSearch at the GEON Portal
• Data delivery HTML, OPeNDAP, OGC (WMS),
  Interoperable
• Basically give me the specific types of data I
  want, for only the specific time and volume I
  specify, and in a way that I can find it quickly
  and easily use it with any application I desire!

4
UNAVCO GEON IDV DevelopmentThe GEON IDV
Website http//geon.unavco.org
Example Below Geodynamic and Tomographic models
on OPeNDAP Server visualization with IDV.
5
Building on the Unidata Integrated Data Viewer
(IDV)

• GEON chose to extend the Unidata IDV as a 3D/4D
  visualization for solid earth science
  applications for a number of reasons.
• The IDV was designed for Amospheric science,
  within a framework of  Cyberinfrastructure.
• The IDV combines visualization with access to
  
   distributed data systems and analysis
  capabilities.
• Powerful 3D-4D visualization
• Embedded mathematical capability using Jython
• Freely available
• Unidata and UNAVCO support and more than
  
   10 person years of
  development
• - Scriptable for server-side automation at GEON
  Portal

6
UNAVCO/GEON Enhancements to Unidatas IDV
The ability to show observations and models for
your domain is essential
Examples

• GPS vectors with error ellipses
• Earthquake focal mechanisms
• Shear wave splitting
• Strain rate
• Customized interface for earth science users

7
About Data
Formats It is desirable to have data in a common
format or at most a few formats, particularly
when dealing with very large 3D/4D models.
Experience so far is that almost no two formats
we get from investigators are the same.
Attributes Need at minimum basic attributes
(x,y,z,t, value(s), uncertainties). Usually can
get this. However, to facilitate integration you
need to employ some attribute conventions and
provide additional information. For example
tomography is typically given as an anomaly. Need
to know reference model for models to be reused
and integrated by others. Georeferencing Often
models are generated only with x,y,z and
information to go to latitude, longitude and
depth is missing. If latitude and longitude are
provided, you may need projections and datums as
well. Boundaries vrs continuum There are two
fundamentally different type of volume
representations. 1) boundaries around volumes of
constant value (e.g. geologic units) and 2) other
gridded approximations of a continuum. Few
visualization programs handle both. Same notion
applies to time. Sampling Data is often
irregularly spaced while gridded models are
uniformly spaced.
8
Why use the NetCDF data format standard?

• A machine-independent binary standard is needed.
  NetCDF provides this as well as rich metadata, a
  data delivery protocol, and API that is used in
  the IDV.
• NetCDF is a flexible container for 1D, 2D, 3D,
  and 4D data attributes and georeferencing
  information are included so that a complete
  NetCDF file is self-describing.
• NetCDF is a mature, supported standard with
  years of development and on-going Unidata
  development and support.
• NetCDF is used by hundreds of science projects,
  and in many applications including the Generic
  Mapping Tool (GMT though in a limited way at the
  moment), Matlab, GRASS 3D. ESRI and Unidata are
  working on ArcGIS support for netCDF.
• There are not a lot of options other than HDF in
  the earth science scientific 3D/4D data/model
  world. NetCDF 4 is including major components of
  HDF5 into a common data model.
• - NetCDF is well supported in the OPeNDAP data
  distribution system.

9
NetCDF Data Files and the IDV
NetCDF files are self-describing and can (and
should) contain information such as units,
dimensions, attributes, and mapping or location
information, and metadata such as data source
attribution (authors and references).
3D Spacial Subsetting (under Data Properties in
the IDV Dashboard)
10
How do you put your data in NetCDF?1. See the
GEON IDV web site2. UNAVCO ASCII-to-NetCDF Data
Conversion Software3. Ask UNAVCO for help
Java application to convert 2d and 3d gridded
data to NetCDF
Data Conversion Software

• Java application to convert 1d and 3d gridded
  data to NetCDF format
• Developed for UNAVCO/GEON by undergraduate
  student research assistant Andy Wahr (Purdue U.)
• Command line interface
• Forms
• Documentation

11
UNAVCO/ GEON Data Access with the IDV
UNAVCO/GEON Data Node OPeNDAP Server Example
Seismic Tomography in NetCDF format WMS
Server Example GPS Seamless Archive
Postgres/PostGIS Database
GEON IDV
THREDDS/OPeNDAP Catalog WMS Catalog HTML
File
Other distributed data servers
Local File
Also -- registration of netCDF files and OPeNDAP
servers embedded in the GEON Portal, discoverable
with the GEONSearch
12

• Acknowledgements

Don Murray and Jeff McWhirter, Unidata Greg
Bensen, UNAVCO and University of Colorado Andy
Wahr, UNAVCO and Purdue Funding NSF EAR and
ATM
13
Data Delivery to the GEON IDV via
OPeNDAP/THREDDS(left) From the IDV Dashboard
select a THREDDS url catalog. The catalog returns
the files on the OPeNDAP server. We serve
primarily netCDF files, and OPeNDAP allows for
server-side data subsetting reducing unnecessary
data transfers.
(Right) Sample tomographic 3D netCDF file
(Berkeley Model) on the UNAVCO GEON OPeNDAP
server.
14
Data Delivery to the GEON IDV via Web Map Service
(WMS)From the IDV Dashboard select a WMS
catalog service. The GetCapabilities REQUEST
returns the catalog of files on the WMS server.
Any WMS server can be accessed.
3D Tomography (Van der Lee, 2005) from OPeNDAP, a
geologic shapefile (from GEON portal), and JPL's
On-Earth WMS satellie image.
15
Data Delivery to the GEON IDV via RSS Feed and
WMS(left) Earthquakes over magnitude 5.0 from
USGS. Map and tabular displays. A click on table
entry centers map on earthquake location.
(Right) Earthquakes and volcanoes plotted with
On-Earth image basemap from JPL Web Map Server
(WMS)
16
Data Delivery to the GEON IDV via Local File or
URL(left) Files can be accessed directly via a
local file system or remotely with HTTP.
(Right) Seismic tomography from local file
system. Isosurfaces and vertical cross sections
generated on demand.
17
Visualization Elements
To be useful for you, a visualization tool must
suit your science domain. What are some earth
science visualization elements we need and how
does the GEON IDV implement them?
18
Visualization Elements - Points

• Locations (x,y,z) of a sample or event such as
  earthquake hypocenters
• Draw with dot, sphere, cross
• Locations and scalar parameter (x,y,z, value)
  such as the magnitude of the earthquake. May want
  to display uncertainty of location and/or value.
• - Draw with dot, sphere, cross
• - Indicate value(s) by color, texture, size,
  intensity
• Location and vectors, such as GPS velocity
• - Draw with vector, but might need conical tip
  for 3D
• - Indicate value with length, vector thickness,
  might use color
• Location and tensor quantity (x,y,z, ) strain
  and stress earthquake source parameter
• - Draw with focal mechanisms anisotropy.
• - Indicate value with color, diameter,
  orientation of faults on focal sphere (really not
  your typical visualization software graphic.
• Boils down to being able to draw arbitrary
  scaled, oriented, colored point symbols. Hard to
  do with GIS. Can be coded into the IDV as needed.

19
Visualization Elements Point Data - Hypocenters
20
Visualization Elements Point Data GPS Plate
Motions
21
Visualization Elements Point Data Focal
mechanisms
22
Visualization Elements Point Data Seismic
Anisotropy
23
Visualization Elements Lines and Polygons

• Lines
• Locations are similar to points but need some
  interpolation to connect the dots. Examples are
  faults mapped on the surface, ray paths, sonde
  tracks, etc.
• Draw with a line or curve
• Lines plus scalar
• Draw with line colored with value, dash line,
  vary thickness
• Draw line with perpendicular variation such as a
  seismic waveform plotted in two or three
  dimensions. Might color in waveform.
• Polygons
• Locations like lines, but enclosed regions of
  constant value. Examples are geologic units of
  constant age or lithology political boundaries.
  Typical GIS shapefile feature.
• Draw polygons with lines
• - Indicate value by area coloring or texture.
  Could raise polygon off map like a histogram. May
  dash polygon boundary or vary thickness.

24
Visualization Elements Lines and Polygons
Lines Polygons
raypaths
shape files
25
Visualization Elements - Surfaces

• Surfaces Maps with relief
• Often surfaces such as topographic contour maps
  with relief are referred to as 3D, they are
  actually 2.5 D surfaces. This is something you
  realize when, for example, you try to grid tripod
  LiDAR data of an overhanging cliff. Here you get
  full 3D with multiple values of Z for each X,Y
  location.
• Draw with lines, polygons (such as triangular
  elements)
• - Indicate value by coloring and texturing the
  polygons can drape raster over surface, show
  relief with perspective and illumination. May use
  transparency.

26
Visualization Elements Raster Files
Raster images can come from photographic imagery,
multi-spectral scanning, interferometric SAR, and
derived products such as WMS and ArcGIS map
servers, gridded data, etc. A significant barrier
to use is geo referencing the image. Geotiff
format allows for this in the raw image
specification, but is not widely used. WMS and
ArcIMS web mapping can also serve up this
information. Otherwise additional files or
metadata are needed. Rasters can be draped over
tomograpy. More recently photographic images have
been applied to full 3D surfaces from tripod
LiDAR scans. Visualization of raster files with
vertical orientation is rarely possible.
Examples from InSAR and LiDAR. Not from IDV yet.
Need to deal with non-standard formats and
rotated images with black background.
27
Visualization Elements Volumes
Volumes can be rendered with isosurfaces 3D
contours of a single value. The surface can be
colored, textured or made transparent to indicate
value or uncertainty. However, work still needs
to be done to effectively show uncertainty in 3D
volume space. IDV also has horizontal and
vertical cross sections, probes, and contouring
on all surfaces.
28
Visualization Data Integration and Exploration
Putting it all together. The first step to true
integration is to remove the barriers to
accessing and importing the data and then
visualization of the data. Her are some examples
of multidimensional data displayed in three
dimensions with the GEON IDV.
Yellowstone (Smith and others) and the
geodynamics of the mantle (McNamara)
29
Visualization of Global Seismic Tomography
Map version of the IDV showing the
Berkeley global shear wave tomography model on a
2 degree grid, Mégnin and Romanowicz, 2000.
Model data from this and other models of the
Reference Earth Model Project are directly
accessed from the UNAVCO/GEON DODS/OPeNDAP,
server
30
Visualization of Mantle Geodynamics (convection)
Mantle Temperature
0.8 T (lower mantle) 0.5 T
(Upper mantle) Whole mantle Convection with
geologic plate motions over 120 million years.
Normalized temperature isosurfaces
shown. McNamara and Zhong (2004)
Lava Lamp analogy? Actually not, the physics is
different.
31
Data Interoperability four linked interactive 3D
globe views satellite photo, seafloor ages,
topography, crustal stain
32
4D Visualization Images from the GEON IDV
170 ma 90 ma
10 ma
Geodynamics- McNamara (convection) Paleomagnetic
s- Schettino, Scotese Paleogeography- Blakely
33
Working in the GEON Portal GEON Search
Search for Resources Current Types ASCII CUAHSI
Data GMT Raster GeoTiff Relational
Database Shapefile Tool WMS Service Web
Service NetCDF Search Constraints Metadata
Relation, Type, Subject, Keyword Spatial
Coverage Temporal Coverage Ontology/Concept
Relation
Example of unconstrained search response
34
Web Service Data AccessThe IRIS Earthquake
Browser
NetCDF data link url
35
Web Service Data AccessThe IRIS Earthquake
Browser data in the IDV
36
MBARI and Yellowstone Demo
MBARI OPeNDAP Server
IDV
IDV
Google Earth
37
Visualization Integration 15 data sets in
Yellowstone
38
Next True stereo 3D Demo
Image source credits Mantle Tomography - Shapiro
and Ritzwoller Megnin and Romanowicz Geodynamics
Model - McNamara and Zhong Global Strain Rate Map
and Plate motions -Kreemer and Holt
39
Later this afternoon ... the GEON IDV
Tutorial http//geon.unavco.org/unavco/tutorial
Contact Stuart Wier wier_at_unavco.org
40
Downloading the IDV from UNIDATA

• IDV User account at www.unidata.ucar.edu for IDV
• Idvuser_at_unidata.ucar.edu (login for download)
• Idvuser (password)
• Note that the GEON IDV is a plugin in the
  Unidata version of the IDV