An open innovation platform for deep time spatio-temporal knowledge-discovery eResearch Sydney 2012 Presentation

Abstract

In proceedings of 6th eResearch Australasia Conference, Sydney, 2012.
Geographic information systems form a core part of Earth Science, allowing the growing repositories of digital geo-data to be integrated and visualised in a unified fashion. These systems cope with the wide variety of spatial data types, each with their own properties and metadata, allowing for a better understanding of how Earth processes operate. A unique requirement for the Earth Sciences is to take into account plate motion and crustal deformation processes acting through time, thus altering the various spatial relationships between datasets. The open-source GPlates (www.gplates.org) software has become a standard tool for this type of analysis, providing the ability to reconstruct various datasets through time interactively by attaching arbitrary data to tectonic plates.
Developed originally as a standlone software, GPlates has gradually grown into an open innovation platform by employing an open-standards-based information model, by implementing extensive data import/export capabilities, developing an emerging python plugin framework and by establishing links to web services. This extensible platform allows users to explore the evolution of the entire Earth system in accordance with past tectonic plate configurations by combining GPlates with a variety of research tools, data and workflows. GPlates applications include deep earth dynamics, tectonics and continental margin reconstruction, evolution of continental stress fields, evolution of river systems, and carbonate reefs, long-term sea- level change, basin evolution, mountain building processes and continental collision, continental paleogeography and the evolution of continental climate and ocean circulation.

Full text

An open innovation platform for
deep time spatio-temporal
knowledge-discovery
Dietmar Müller, John Cannon, Thomas
Landgrebe and Michael Chin
The University of Sydney
Tuesday, 7 January 14

Paleogeographic Information Systems
•Geographic information systems form a core part of Earth Science
•Allow the growing repositories of digital geo-data to be integrated
and visualised in a unified fashion.
•Systems cope with the wide variety of spatial data types, each with
their own properties and metadata, allowing for a better
understanding of how Earth processes operate.
•A unique requirement for the Earth Sciences is to take into account
plate motion and crustal deformation processes acting through time,
thus altering the various spatial relationships between datasets.
•The open-source GPlates (www.gplates.org) infrastructure has
become the leading, standard tool for this type of analysis, providing
the ability to reconstruct various datasets through time by attaching
arbitrary multi-dimensional data to tectonic plates.
Tuesday, 7 January 14

GPlates Open Innovation Platform
•Developed originally as a standlone software, GPlates has gradually
grown into an open innovation platform (AuScope, ARC LF)
•Open-standards-based information model, implementing
extensive data import/export capabilities, developing an
emerging python plugin framework and by establishing links to
web services.
•Extensible platform allows users to explore the evolution of the entire
Earth system in accordance with past tectonic plate configurations by
combining GPlates with a variety of research tools, data and
workflows.
•GPlates applications include deep earth dynamics, tectonics and
continental margin reconstruction, evolution of continental stress
fields, evolution of river systems and carbonate reefs, long-term sea-
level change, basin evolution, mountain building processes, continental
paleogeography, and the evolution of climate and ocean circulation.
Tuesday, 7 January 14

Solid Earth e-geoscience drivers
Understanding the evolution of the Australian continent
its internal boundaries, and its offshore territories, in a
global and a deep Earth context
Understanding the geodynamic context of mineral and
energy systems in their original geological settings
Need to amalgamate an enormous diversity of
data for joint analysis and modelling, towards
construction of predictive models
Tuesday, 7 January 14

In 2010 the Academy hosted its annual High Flyers Think Tank on
Searching the Deep Earth: The Future of Australian Resource Discovery
and Utilisation.
A valuable opportunity for 60 of Australia’s leading early and mid career
researchers to identify and propose new directions for Australian minerals
exploration research.
Participants identified new approaches, technologies, data management
systems, and policy innovations to facilitate the science necessary to
deliver a better understanding of the deep earth and ultimately help to
maintain mining productivity into the future.
Tuesday, 7 January 14

Lambert, 2010
Tuesday, 7 January 14

Tuesday, 7 January 14

Tuesday, 7 January 14

Tuesday, 7 January 14

Tuesday, 7 January 14

11
SEARCHING THE DEEP EARTH A VISION FOR EXPLORATION GEOSCIENCE IN AUSTRALIA
•National Cover Map
•National Map of the Deep Crust & Mantle
•National 4-D Metallogenic Map
•National Distal Footprints Map
•National Researcher Network
•National Education Outreach & Transfer
Program
Searching the Deep Earth: The future of
Australian resource discovery & utilisation
Tuesday, 7 January 14

GPlates history and specs
•10+ years of development, starting in 2003
•Collaborative development includes Caltech, NGU and TUM
•Open-source and platform-independent
•Information model and file format based on Geographic
Markup Language (GML, ISO standard)
•Interoperability with ArcGIS and QGIS (shapefile I/O)
•Interoperability with netcdf grids (raster data standard)
•Interoperability with open source data-mining software
(Orange) underway
•Linking to web feature services (WFS) underway
•GPlates service on NCI’s MASSIVE system
Tuesday, 7 January 14

Workflows for linking tectonic models via GPlates to HPC codes:
CitcomS, Terra, Boundary Element Code (BEM), Slim3D,
Rhea
Additional links to many other codes established or underway
(Underworld, Tellus, Madagascar, GoCAD, ArcGIS, Quantum GIS,
Generic Mapping Tools, Orange data mining)
Links to various web feature services and online databases
underway (e.g. paleobiology, paleomag, GA data); future: VGL data
Python plugin infrastructure under development will enable easy
interoperability with other codes such as escript and data sources,
EarthScope interoperability
Users can build regional or global models, import their own
data and digitise features in a “Paleo-GIS” environment
GPlates workflows and extensions
Tuesday, 7 January 14

•> 30,000 downloads
•Users in 137 countries
•Anyone who studies the Earth back
through time
•Geologists and Geophysicists, Plate
modellers, Geodynamicists, Resource
Exploration
•Paleogeographers,
Paleoceanographers, Paleobotanists
•Educational Tool
Who uses GPlates ?
Tuesday, 7 January 14

GPlates Interoperability
Tuesday, 7 January 14

Spatio-temporal data-mining
Large igneous provinces
at 820Ma (no
reconstruction)
Large igneous provinces at 820Ma
(Rodinia reconstruction)
• Data mining is about finding patterns in complex data
• In exploration we need to consider Earth’s plate tectonic
history to reveal them – in this case the plate alignment at 820
Ma reveals a single, giant magmatic event
Gairdner LIP
Aksu LIP
Suxiong-Xiaofeng LIP
Tuesday, 7 January 14

Data-mining – SMH, 12 May 2012: “Data miner is
the hottest new job you've never heard of”
There's another mining boom
you may have missed. It too
involves paying young people six-
figure salaries in their first jobs,
and exploring deeper for
resources which may have been
previously overlooked. But it's
not about driving trucks or
digging holes. It's about building
algorithms and crunching facts
and numbers. It's mining for data.
''The whole place is big data mad.
Industries like banking, insurance,
and increasingly pharmaceuticals
are competing on the back of
predictive models that get built
by mining data.''
Tuesday, 7 January 14

Data-mining library options
Library
Functionality
Flexibilty/
Extensibility
Compatibility
Usability
Licensing/
platforms
PRTools
Excellent
Excellent
No - Matlab
Experts only
Library open but
Matlab commercial
AiLab Orange
Good
Excellent – has
plugin infrastructure,
plotting tools, scipy,
numpy
Yes - python
Visual programming
environment to
abstract complexity
Open source, cross-
platform
Weka
Very good
Excellent – has
plugin infrastructure
No - Java
Visual programming
environment
Open source, Java
i.e. cross platform
RapidMiner
Excellent
Excellent – but
biased towards
statistics, not signal
processing
No - Java
Visual programming
environment
Open source and
commercial options
MDP
Limited
Good
Yes - python
Experts only
Open source, cross
platform
Waffles
Limited
Limited
Yes – C++
Experts only
Open source, cross
platform
Tuesday, 7 January 14

AILab Orange
Visual programming canvas – can
abstract complexity.
Several data-mining plugins
including data manipulation,
supervised/unsupervised
classification, plotting tools etc
Make your own “widgets” by
writing Python scripts
e.g. GPlatesPalaeoAssociation
plugin has widgets to analyse a
timeseries
Python Matplotlib is used to create
useful plots e.g. scatterplots
Tuesday, 7 January 14

!"#$%&'()*+%(,-)%.)/0-$"#+'#1)-'23+'456)#75")895"-)5$)#+)
:;<<=>?)*"'@A$5")B+05)#"5#-)#"5)("#$%&'()B+%(,-)$A5)C%"$A1)
D%0$A)#&6)E5-$)/0-$"#+'#&)!"#$%&-):C/!1)D/!)#&6)E/!)
"5-35(FG5+9>?)H#",5")B+05)#"5#-)(%23"'-5)I"%$5"%J%'()#&6)
9%0&@5")0&'$-1)-0B6'G'656)$%)#++%K)$A5-5)0&'$-)$%)B5)
#L#(A56)$%)6'M5"5&$)("#$%&-)'&)$A5)#+$5"&#FG5)
"5(%&-$"0(F%&)2%65+-?)8*)N)80-@"#G5)*+%(,?)
!"0-$)$%)$A5)5#-$)%.)$A5)O#-2#&)P'&5).%"256)B9)
IA#&5"%J%'()#(("5F%&?):BQ6>)6'M5"5&$)3+#$5)
"5(%&-$"0(F%&)-(5&#"'%-)3"%3%-56).%")/0-$"#+'#)-'&(5)$A5)
5#"+9)85-%3"%$5"%J%'():R;=SS)8#>?)
:B>)#75")P')#&6)TG#&-):US;;>1)#)VSW)"%$#F%&)%.)C/!X)
:(>)Y'+5-)5$)#+):USSV>1)ZUW)"%$#F%&)%.)D/!X)
:6>)#75")[5&-%&)5$)#+):US;;>1)$"#&-+#F%&)%.)C/!)"5+#FG5)
$%)E/!)#&6)D/!?
!"# !"# !"#
$"# $"#
$"#
%"# %"#
%"#
%"#
!"#
$"#
&'
O#-2#&)P'&5
/
* ! H
Tuesday, 7 January 14

Spatio-temporal data miner
•Can help to derive a candidate plate motion model
Gravity
Magnetics
Mafic/ultramafic events
Radiometrics
Develo
p
Tuesday, 7 January 14

&( &(
&(
#)
#)
#)
*+,-./0)1/02342546/-7+89/:1-+947+6;/<-=4>-
")<3>+6;+?-@+3+-+>/-3+A/9-=>47-3B/-")<3>+6;+-
7+89/:1-1475;6+:49-+90-+-C+>;+D6/-;916;9+:49-
>/0)1:4923423B/2546/-+556;/0-34->/74C/-
;916;9+:4920/5/90/91/-4=-3B/-<B+5/-4=-;90)1/0-
+947+6;/<-*&;66;8+9-EFGGH-5/><-1477?,-
*D20,-&+89/:1-0+3+-=>47-*",->/149<3>)13/0-)<;98-
I-1+90;0+3/-56+3/-149J8)>+:49<-<B4K9-;9-J8)>/-LM-
*+,-+N/>-O;-+90-PC+9<-*EFGG,M-
*D,-Q;6/<-/3-+6-*EFFI,M-
*1,-+N/>-R/9<49-/3-+6-*EFGG,?-!B;3/-4)36;9/<-<B4K-
3B/-#)>9+749+-S>4C;91/-*6+D/66/0-#),-+90-&4)93-
(<+-'641A-*6+D/66/0-&(,?
/ * !
/
Tuesday, 7 January 14

I+#$5)O5($%&'()"5(%&-$"0(F%&)'&)YI+#$5-)#$)\]S)8#1):#>)'-)B#-56)%&)$A5)B+%(,)654&'F%&-)#&6)
3%+5-)%.)"%$#F%&)%.)P')5$)#+):USS]>X):B>)0-5-)$A5)-#25)B+%(,)654&'F%&-)B0$)$A5)3%+5-)%.)"%$#F%&)
%.)TG#&-):USS<>?)OA5)6#$#)#"5)$A5)K%"+6)2#@&5F()#&%2#+9)2#3):T8/YU1)8#0-)5$)#+)USS<>?)
P#0"5&F#
Y"55&+#&6
*#+F(#
/0-$"#+'#
D'B5"'#
D?)!A'&#
/2#J%&'#
!%&@%
E?)/."'(#
^#+#A#"'
D#%)
_"#&('-(
%
8#K-%&
C)!A'&# P#0"5&F#
Y"55&+#&6
*#+F(# /0-$"#+'#
D'B5"'#
/2#J%&'#
!%&@%
E?)/."'(#
^#+#A#"'
D#%)
_"#&('-(
%
8#K-%&
C)!A'&#
/2#J%&'#
`&6'#
a'%)65)
+#)
I+#$#
/ *
Testing of alternative supercontinent
reconstruction hypotheses: Rodinia
Tuesday, 7 January 14

Spatio-temporal data miner
•Analyse data from many crustal fragments, eg
those formed between 1.7 and 1.5 Ga (share
event chronology segments)
Magnetics
Gravity
Radiometrics,
Potassium
Radiometrics,
Thorium
Radiometrics,
Uranium
Tuesday, 7 January 14

Constrain reconstructions
beyond the Phanerozoic
Palaeo-alignments
Age-tag mineral systems
Prediction
4-D Metallogenic Maps
Australia
Laurentia
South
China
Reconstructed images comprise
• global 2 minute grid
• higher resolution public domain grids for North
America and Australia
Greater level of detail for Australia is revealed
when we zoom in
Tuesday, 7 January 14

GPLates1.3 (release in April 2013): Volume
visualisation
Two central isosurfaces with isovalues isovalue1 and isovalue2.
Each with its own deviation window (symmetric or asymmetric).
Mantle
Temperature
Shepard et al., 2012 EPSL
Tuesday, 7 January 14

Two central isosurfaces with isovalues isovalue1 and isovalue2.
Each with its own deviation window (symmetric or asymmetric).
Mantle
Temperature
Shepard et al., 2012 EPSL
GPLates1.3 (release in April 2013): Volume
visualisation
Tuesday, 7 January 14

GPlates
Online
Resources
Step-by-Step Tutorials,
Reference Manual
Tuesday, 7 January 14

Key GPlates papers
Boyden, J.A., Müller, R.D., Gurnis, M., Torsvik, T.H., Clark, J.A., Turner, M., Ivey-Law, H.,
Watson, R.J. and Cannon, J.S., 2011, Next-generation plate-tectonic
reconstructions using GPlates, in: Geoinformatics: Cyberinfrastructure for the
Solid Earth Sciences, Keller G.R. and Baru, C., eds., Cambridge University Press, p.
95-114.
Williams, S., Müller, R.D., Landgrebe, T. C.W., Whittaker, J.M., 2012, An open-source
software environment for visualizing and refining plate tectonic
reconstructions using high resolution geological and geophysical data
sets, GSA Today, 22, no. 4/5.
Gurnis, M., Turner, M., Zahirovic, S., DiCaprio, L., Spasojevich, S., Müller, R.D., Boyden,
J., Manea, V.C., Bower, D. and Zahirovich, S., 2012, Plate Reconstructions with
Continuously Closing Plates, Computers and Geosciences, 38, 35-42.
Seton, M., Müller R.D., Zahirovic, S., Gaina, C., Torsvik, T., Shephard, G.E., Talsma, A.,
Gurnis, M., Turner, M., and Chandler, M., 2012, Global continental and ocean
basin reconstructions since 200 Ma, Earth Science Reviews, 113, 212–270.
Qin, X., Müller, R.D., Cannon, J., Landgrebe, T.C.W., Heine, C., Watson, R.J., and Turner,
M., 2012, The GPlates Geological Information Model and Markup
Language, Geosci. Instrum. Method. Data Syst. Discuss., 2, 1–63, www.geosci-instrum-
method-data-syst-discuss.net/2/1/2012/ doi:10.5194/gid-2-1-2012
downloadable at www.earthbyte.org
its offshore territories, and its offshore territories, and
Tuesday, 7 January 14

NeCTAR/ANDS #nadojo
competition winner 2012
•Michael Chin from the Earthbyte group won this year's
NeCTAR/ANDS e-research 2012 #nadojo competition
•Demonstration of a new web-enabled GPlates prototype on
the NECTAR/ANDS eResearch computing infrastructure
•Porting some of the new tectonic plate reconstruction
capabilities to the internet
•The web-enabled GPlates infrastructure won on the basis of
its novelty and future potential
•A judge was quoted as saying: "...being able to look at tectonic
plates and how they move over 140Ma years ago brought me
back to that child-like awe of wonder and amazement, like the
first time you look up at the stars and realise how many (light?)
years away they are..."
Tuesday, 7 January 14

Tuesday, 7 January 14

Tuesday, 7 January 14