Upload
chastity-tucker
View
214
Download
0
Tags:
Embed Size (px)
Citation preview
GIS-based visualization and map server efforts in
support of marine fisheries and ecosystem management
Tiffany C. Vance (AFSC) and Christopher Moore (PMEL)
Nazila Merati PMELJason Fabritz OMAOHal Mofjeld PMEL
VRML based visualizations for the Cordell Bank NMS
Using ArcIMS map servers for intra-layer calculations
Java/Java3D and ArcGIS Engine as a framework for a “scientific GIS”
Introduction Visualizations of spatially complicated
datasets are used to enable scientists to understand complex physical and biological processes.
These geo-visualizations are also becoming a way to disseminate the data as a coherent package.
Rather than distributing discrete datasets, • a project can disseminate a view of the data • the recipient has the ability to move through the
data• can add and remove layers • can query to datasets at specific three-
dimensional locations.
VRML based visualizations for the Cordell Bank NMS
Using ArcIMS map servers for intra-layer calculations
Java/Java3D and ArcGIS Engine as a framework for a “scientific GIS”
Project goals
To create interactive visualizations of GIS data for the Sanctuary
To enable viewers to see the Sanctuary as a volume, not a flat map
To test distribution of the visualizations
To test integration of GIS displays and true 3D visualizations
Cordell BankCordell Bank Marine
Sanctuary is a 526-square mile sanctuary located 50 miles northwest of San Francisco. The Sanctuary encompasses Cordell Bank - a pinnacle rising from the seafloor to within 120 feet of the sea surface - and the surrounding waters.
Datasets for the Sanctuary
Bathymetry Physical
characteristics- CTD Hydroacoustic surveryBottom type data SST imagesCoastlines and
boundaries
Visualization of the Sanctuary
Standard tools and plug-ins within browsers (VRML players, animation players such as RealPlayer, javascript tools) to enable users to manipulate the visualizations.
Viewing a visualization
VRML generation - fencelines
Isosurfaces, plumes and vertical fenceline plots created using EVS-Pro. EVS-Pro allows for3-D kriging and fenceline plots
Fenceline of towed instrument data
VRML generation - isosurfaces
3D temperature plumes and isosurfaces created in EVS-Pro are exported and combined with the VRML 2.0 output from ArcScene
(8 degree temperature isosurfaceand CTD cast positions)
Viewing visualizations User loads the visualization into
a VRML-aware web browser Coastline, bathymetry and
topography data in the VRML window.
3-D navigation control in the VRML window
Can load, view and animate data as the scene is rotated and scaled.
Radio-button choices are given for dataset choices
Animation controls appear as time dependent data are loaded.
VRML based visualizations for the Cordell Bank NMS
Using ArcIMS map servers for intra-layer calculations
Java/Java3D and ArcGIS Engine as a framework for a “scientific GIS”
Project goals To create a series of tools to allow
user defined intra and inter-layer calculations and comparisons within the framework of ArcIMS
To allow PMEL to be able to calculate the population at risk from tsunamis
To allow NMML and AFSC to calculate biophysical measures
Background Internet map servers (IMS) are used to
• disseminate information• allow users to perform queries• to extract information • to serve data
All line offices in NOAA are using IMS applications to serve data
A drawback of off-the-shelf map servers is that one cannot do on-the-fly calculations on layers, or between layers
WebMapCalculator architecture
ArcIMS/JSP
Application
HTTP PostWebsite
Server Side calculator
File Access
ShapeFiles
Java Servlet
Application
Calculated Results
Using ArcIMS 4.0.1 on Solaris with JAVA JDK 1.4, Running Image Server, Feature Server and Extract Server
Path to Files
Include Shapefiles in ArcIMS Map Workspace
Demonstration project - tsunami modeling in Puget Sound
PMEL’s Tsunami Inundation Mapping Effort (TIME)
Products for use by emergency managers Involves ingesting data from
• municipalities• NOAA• model output • observational data
Data products are • maps - static and live • data reports produced using GIS analysis
TIME data
Data are disseminated as ArcView projects. Layers include • inundation fields • census products • run up model
results • animations.
Distributed to emergency managers via CD
WebMapCalculator for TIME
Input: Gridded wave height
data from a tsunami model
Population of Seattle area by day and night
Elevation data Output: A polygon that shows
resulting at-risk population
Output from the IMS Inundation IMS shows
users • inundation results
from model• maximum velocities• day/night populations• natural hazards• shoreline data
Data sources have metadata associated with the layers
Toolkits to be added
RACEBASE dataset of trawl survey data - intra-layer calculations between fisheries datasets and physical oceanography datasets such as water temperature or salinity
NMML - ability to query tracked mammal results to determine swimming speeds, distanced traveled
VRML based visualizations for the Cordell Bank NMS
Using ArcIMS map servers for intra-layer calculations
Java/Java3D and ArcGIS Engine as a framework for a “scientific GIS”
Project Goals To extend the capabilities of ArcGIS to
form the foundation of a “scientific GIS” for fisheries oceanography
To integrate existing oceanographic analytical tools with ArcGIS
To take advantage of visualization tools such as VRML and Java3D to provide truly three-dimensional visualizations
Programming options ArcObjects/VB - limited to single
platform, limitations of VB ArcGIS Engine - platform
independent, cost? Open source GIS tools such as
GRASS, MapServer, PostGIS, GeoTools and VisAD - documentation/support
System Diagram
Algorithms
UNESCO routines for water properties Oceanographic Analyst (ArcView 3.2)
http://www.absc.usgs.gov/glba/gistools/
Matlab tools - SEA-MAT package http://woodshole.er.usgs.gov/operations/sea-mat/
USGS sedx package http://woodshole.er.usgs.gov/staffpages/csherwood/sedx_equations/sedxinfo.html
VTK toolkit - for volume analysis http://public.kitware.com/VTK/
Test Case - Mixed layer depth
www.fd.ntou.edu.tw/5CTemperature201025.doc
The depth to which water is well mixed. This has ramifications for fish and planktonic organisms, also for nutrients.
Surface layer sits above the thermocline. Defined as the layer where the temperature is within 0.5° of the average surface temperature or where the potential density is within 0.125 of the surface average
Conductivity-temperature-depth (CTD) data
Java test case MLD algorithm from VB to Java GeoTools toolkit shapefile reader (Java)
used to read shapefile Created a new application in Java to
calculate the MLD and output a VTK OpenGL window
VTK wrapped in Java Can also display MLD shapefile created in
ArcGIS version
3-D Visualization at PMEL
1. Perspective:1. Perspective:
2. Relative Motion:2. Relative Motion:
3. Stereo:3. Stereo:
Why 3-D?
1. Perspective
High frequency spikes in the bathymetry data are obvious in the 3D plot (right) and are obscured in the 2D plot above.
Calculations of bathymetry gradients to identify regions of internal tide generation would be impacted by these spikes in the bathymetry data.
Bathymetry in Astoria Canyon offshore from the Columbia River outflow in Washington State, in 2D and 3D.
2. Relative Motion (Interaction)
• The ability to judge an object’s distance through the use of relative motion
3. Stereographic Virtual Reality
Fish larvae in a canyon
Stereo gives the scientist true depth perception
Stereo
Stereo
Mono
Mono
Ocean currents
21
ImmersaDesk
A Next Generation Internet (NGI) Testbed
The ImmersaDesk:
• 4’ x 5’ rear projecting screen
• near immersive• 1024 x 768 x 96 Hz• driven by SGI Onyx2
• Two R12000 Processors• 250 MHz• Infinite Reality Graphics
GeoWall
PC-driven projection systemStereoCommodity graphics cardsInexpensiveNOAA-Tech
Hostcomputer
polarizing filter
Projector(R
frame)
Projector(L
frame)
polarizing filter
Polarization-preserving
screen
Supports *any* stereo-equipped software:
vis5d, visAD, stereo VRML viewers, etc.
The GeoWall Approach
Problem:
We’re pushing the computational limits with our models. Even high-end graphics cards aren’t up to the challenge
Let’s look at a real-world example…
PMEL scientist models Gulf of Alaska
Uses NCSA supercomputer cluster to modelLarge domain (540 x 320 x 32) = 5.5 million pointsGenerating files on the order of a terabyte (1000 gigs)
Our models aren’t just run on a linux cluster, they are runon several clusters, connected using Grid technology:
TeraGrid - connecting heterogeneous clusters Myrinet Myrinet
Chicago & LA DTF Core Switch/Routers
SunServer
Federation
7.8 TFPower4
1 TFItanium2
Fibre Channel Fibre Channel
2 TF Itanium29.2 TF Madison
0.5 TF Itanium290 TB
1.5 TF Itanium2/Madison20 TB
DatawulfIA-32
SDSC NCSA
Caltech ArgonneQuadrics
PSC
6TF Alpha EV681.1 TF Alpha EV7
300 TB 300 TB 160 TB
VisAD – Java-based Graphics Tool
VisAD uses Java3D to render 3-D scenes Java provides a Remote Method Invocation (RMI) that allows data to be rendered at each “node” of a cluster, and then stitched together at the client.
host(also a node)
client(PC)
Clusternodes
RMI
internet
*VisAD and RMI framework for parallel rendering by Bill Hibbard: http://www.ssec.wisc.edu/~billh/visad.html
VisAD test program
Viz Clusters: Distributed Rendering with the GeoWall2
• Developed at EVL & SDSC, SCRIPPS
• 15 LCD screens in 3x5 array driven
• by small Linux cluster
• Total resolution: 8000x3600
• Video compositing allows each node to render from distributed file - up to 38 Terabytes of data on the screen!
• Software: JuxtaView, ParaView
• Scalable: Personal GW is 2x2
Future activities
Framework for 3D modeling of environmental factors
Use of Java to handle temporal analyses
Other graphics outputs Integration with ArcIMS site
This work was funded by NOAA’s HPCC program (http://www.cio.noaa.gov/hpcc/) and the Sanctuaries Program (www.sanctuaries.nos.noaa.gov).
For more information about the Pacific Marine Environmental Laboratory's visualization efforts, please visit the PMEL visualization page at http://www.pmel.noaa.gov/vrml/3DViz.html and http://www.pmel.noaa.gov/visualization/.
Questions?