Potential for Nearrealtime Dataflow Monitoring with GeoWall

1
Potential for Near-real-time Dataflow Monitoring
with GeoWall

• Dr. Kent Lindquist
• Lindquist Consulting, Inc.
• May 10, 2004

2
ROADNet Collaboration

• ROADNet Team
• Atul Nayak, Todd Hansen, Rob Newman,

3
Real-time Data Observatories

• ROADNet Live, working testbed
• Sensor grid
• Data grid
• Large Data Volume (multiple GB/day)
• Many Hundreds of Sensors
• Geographically Distributed
• (of course! Its earth monitoring)
• Fundamentally a 3D (4D) problem
• Multidisciplinary

4
Challenges

• How do you
• Make sure the sensor network is all working
• Get intuitive overview of whats happening

5
Application Domains

• Seismic Monitoring
• HF Radar Ocean-current Monitoring
• Remote cameras
• Satellite data
• Many others water quality, stream level, fuel
  moisture, fuel load, NMEA, ADCP, shipboard GPS
  and Gyroscopic data, Land-based GPS, Weather,
  etc.
• Plus Network engineering metrics
• Dataflow
• Dynamic Network Topology

6
Real-time Monitoring Challenges

• Seismic Network State of Health
• Which stations are operating
• Which are behind (time lag greater than expected)
• Which have gaps, glitches, etc.
• Which have stopped sending data
• What patterns are there in any problems (gtcause)
• Past approaches
• 2D scrolling waveform displays
• Monitoring scripts
• Manual data-scanning
• Daily summary emails
• Summary web pages, gap reports

7
2D Seismic Monitoring

• Near-real-time packetized waveform acquisition
• Includes energy detection, network triggering,
  phases
• Hard with 400 stations

8
Sample seismic station map
9
Real-time Monitoring Opportunities

• Four dimensional character of monitored wavefield
• Better overview patterns emerge, telemetry
  problems etc.
• Wavefield propagation
• Visual integration of complex, time-dependent
  events
• Single-station detections
• Network triggers
• Earthquake Locations
• Earthquake Vital statisticssize etc.
• Earthquake effects QuakeMaps
• Information density and clarity of relationships
  in 4D view allows human pattern recognition to
  work at its best

10
Latest Auto-detected Earthquakes

• Value added from processing
• Earthquakes automatically detected within minutes
  by automated system

11
Virtual Reality Earthquake View
12
Earthquake Ground-motion estimation
Quick script to show Qgrids out of a database
Peak Ground Acceleration Trinetsm_es99 delegate
May 31, 2003 1712 GMT ML 3.14 Max 0.076
g Lat 32.6513 Lon -116.7682
13
Earthquake Ground-motion Estimation

• May 12, 2003
• 2235 GMT
• ML 2.95
• Max 0.084 g
• Lat 33.2710
• Lon -116.0527

dbdbopen('ssn','r') dbdblookup(db,'','qgrid','o
rid','24') cggcggrid(dbfilename(db)) x,y,zcg
grid_getmesh(cgg) mysurfc(x,y,z)
14
Joint Image / Seismic Monitoring
Coronado Bridge Monitoring Experiment May 15,
2002 Structural-health camera For bridge
bearing Seismic strong-motion instrument Array
of distributed cameras needs easy overview
15
Near-real-time Time-Lapse Movies

• Geographically Distributed Cameras
• Relational Database of source images
• Need GeoWall Stereo N-R-T movies!

16
Current GeoWall Efforts

• Wiggleview
• Live seismic data on GeoWall
• Live Seismoglyphs
• Near-real-time geo-located image display

17
Wiggleview Seismogram Display
18
Wiggleview Seismograms
19
Seismoglyphs
20
Geo-referenced Images
21
Pinon Flats Observatory Image
22
Summary

• GeoWall can be a powerful resource for
  Geophysical Data Observatories
• Inherently 4D visualization problem
• Overview of Network State of Health not available
  with 2D applications
• Fast, Intuitive Overview of Near-real-time Earth
  Events information density integrates the
  picture more than 2D media can achieve
• Suited for multidisciplinary study (ROADNet)