A Discussion of the Visualization Needs of the Community - PowerPoint PPT Presentation

1 / 34
About This Presentation
Title:

A Discussion of the Visualization Needs of the Community

Description:

... earth (global) and flat earth (projected) ... Support many map projections and earth ellipsoids ... Unidata Integrated Data Viewer (from the IDV User's Guide) ... – PowerPoint PPT presentation

Number of Views:22
Avg rating:3.0/5.0
Slides: 35
Provided by: aliso72
Category:

less

Transcript and Presenter's Notes

Title: A Discussion of the Visualization Needs of the Community


1
A Discussion of the Visualization Needs of the
Community Where do we go Next?
  • Alison M. Eyth
  • Carolina Environmental Program
  • University of North Carolina at Chapel Hill
  • October 29, 2003

2
PAVE Smoothed Tile Plot Animation(courtesy RWDI)
3
Currently Used Visualization Tools
  • PAVE
  • free to use, recently made open source
  • tile plots and integrated time series plots
  • supports 2D and 3D temporal varying gridded data
    and point observations
  • supports animation and formulas
  • GIS
  • Free GRASS, ArcExplorer, OpenMap,
  • License fees MapInfo, ArcView, ArcGIS,
    MapViewer,
  • regular and thematic maps of data in points,
    lines, polygonse.g. geopolitical boundaries,
    roads, water bodies
  • supports multiple layers of different data types
  • Free tools Vis5D, NCAR Graphics, OpenDX, GrADS,
    FAST,
  • Commercial tools SAS, Matlab, S-PLUS,

4
PAVE Plotting Relative Difference (courtesy NC
DNR)
5
PAVE Plotting Gridded Data and Observations
(courtesy NC DNR)
6
Some PAVE Limitations
  • Graphics are written in X-Windows / Motif (not
    good for PCs)
  • Cannot plot data in lines or polygons only
    rectangular grid cells and limited capability for
    point observations
  • Cannot incorporate GIS data such as population
    density or other interesting data sets into plots
  • Cannot be executed over the Web
  • Cannot animate over x, y, or z (only time)
  • Only a few maps provided (states, counties,
    roads, rivers), with limited rendering styles
  • Lack of support for 3D plots or contour plots
    with lines
  • Must be compiled separately for each platform
  • Difficult to maintain

7
GIS Example ESRI ArcExplorer(Layers, Colored
Thematic Map)
8
Color-based Thematic Map of Polygons(Courtesy
LADCO)
9
Size-based Thematic Map of Point Sources
(courtesy LADCO)
10
GIS-style Annotated Thematic Map of Gridded
Data(courtesy RWDI)
11
Limitations of GIS
  • Many are expensive for end users
  • But GRASS and ArcExplorer are free
  • Can be complicated to use
  • Little support for time-varying or 3D model data
  • Would need to be customized to read standard AQM
    data formats
  • Limited (any?) support for formulas
  • Animation capabilities very limited
  • Few cross-platform options (usually they are
    either PC or Unix)

12
Hourly Stacked Bar Chart(courtesy LADCO)
13
Bar Chart Plus Line(courtesy LADCO)
14
Vis5D Vertical Cross Sections, Terrain, Isosurface
15
Vis5D Spreadsheet
16
Can we Create The Ultimate Visualization Tool
  • Combine the good elements from existing commonly
    used visualization tools into one ultimate
    visualization tool
  • Goal Meet almost all the visualization needs
    within a single tool
  • Many aspects of requirements for the tool
  • Data
  • Display
  • Functional
  • System

17
Data Requirements for The Ultimate
Visualization Tool
  • Read formats common to meteorological and air
    quality models
  • direct AQM inputs and outputs (2D and 3D
    temporally varying gridded data, gridded boundary
    files)
  • emission inventories (temporally varying point,
    line, and polygon data with attributes)
  • Read inputs, outputs and observational data for
    other types of environmental models
  • multimedia models, water quality models, plume
    models, irregular grid air quality models,
  • Read and display maps/data from GIS files (e.g.
    Shapefiles)
  • Geopolitical boundaries, water bodies, road
    networks, cities, land use
  • Read DTED, satellite images, radar data, met and
    air quality observational data, flight paths with
    observations
  • Read local and remote files
  • Read very large data sets
  • Understand 3-D and time varying data

18
Display Requirements forThe Ultimate
Visualization Tool
  • Present diverse types of data in an integrated
    display (2D 3D gridded, point, line, polygon,
    satellite, DTED, )
  • Have round earth (global) and flat earth
    (projected) displays
  • Standard and thematic maps for points, lines, and
    polygons
  • Satellite, radar, (and other) image displays
  • 3D displays e.g. Topography, multi-colored
    isosurfaces
  • Line contours and color filled contours
  • Color filled grid cells, continuously shaded
    gridded data
  • Met displays Wind vectors, wind barbs,
    streamlines, Skew-t plots, met observations
  • Integrated access to time series plots, bar
    charts, histograms, and other 2D charts from
    map-based plots
  • Support multiple related display windows (with
    concurrent animation)
  • Support for transparent colors (helps w/
    integrated display)

19
System Requirements for The Ultimate
Visualization Tool
  • Support both batch and interactive modes
  • Run over the web and on desktops
  • Run on Windows and Unix systems
  • Manage memory appropriately to analyze very large
    data sets
  • Save plots and animations to image files
  • Facilitate creation of web pages for its outputs
  • Print plots
  • Easy to compile
  • Easy to extend (e.g. to new data displays, data
    formats)
  • Easy / intuitive to use

20
Functional Requirements forThe Ultimate
Visualization Tool
  • Animate through time and space (e.g. x, y, z,
    arbitrary plane)
  • Select and plot subsets of data sets
  • Probing to inspect data values / attributes (e.g.
    in tables)
  • View slices of 4D datasets using various 2D
    displays (e.g. time series plots, x-t plots, bar
    charts, tile plots)
  • Compute and plot results of formulas
  • Perform data interpolation (e.g. point data onto
    a grid) plot
  • Compute useful statistics about data
  • Customizable plots (e.g. legend, colors, header,
    footer)
  • Zoom, pan, rotate

21
GIS-like Functional Requirements forThe
Ultimate Visualization Tool
  • Support many map projections and earth ellipsoids
  • Save / reload project so you can restart where
    you left off
  • Provide overview map for zooming panning
  • Query to find data objects that meet a criteria
    and analyze the attributes of or operate on
    resulting objects
  • e.g. find all point sources emitting gt 100 tons
    NOx / year orfind all grid cells with max ozone
    gt 120ppb) and show on map or in a table
  • Measure distances / areas along map and show
    scale of map
  • Show coordinates of mouse pointer in lat-lon and
    projected coordinates
  • Show data sets only at scales appropriate to the
    data set
  • Label features with attribute data
  • Show pop up tool tips based-on attribute data
  • Find addresses on a map

22
GIST QA of Point Source Inventory
23
Geographic Information System Tool (GIST)
  • Prototype developed at MCNC to explore features
    not available in PAVE
  • Integrated display of point, line, and polygon
    data in layers
  • Reads Shapefiles, extensible to read other data
    formats
  • Creates standard and thematic maps
  • Supports common map projection types
  • Customizable plots wrt colors, shapes, line
    styles
  • Probing to inspect attributes of data objects
  • Querying of attribute data then map or see table
    of matches
  • Show coordinates of mouse pointer
  • Overview map, zoom, pan
  • Java-based runs over web, on Windows / Unix
    without recompilation
  • Easy to use
  • Interactive mode only
  • Had planned to support animated thematic maps and
    reading model data, but didnt get to that point

24
GIST Query to Find Objects that Meet a Criteria,
Show Results on Map or in Table
25
Visualizing Grids with MIMS Grid Family GUI
26
MIMS Grid Family GUI
  • Define and visualize a set of related grids
  • View grids in conjunction with shapefiles of
    related data (e.g. counties, rivers, point
    sources)
  • Support for creating fine grids nested within a
    coarse grid
  • Supported map projection types are
    Latitude-Longitude, UTM, and Lambert conformal
  • Grids defined by four dimensions lower left x
    y, upper right x y, dx dy, number of cells x
    y
  • Specify three dimensions and compute fourth
  • Can specify lower left upper right corners with
    mouse
  • Java-based. Currently available with MIMS (see
    http//www.epa.gov/asmdnerl/mims)
  • Write grids to / read grids from GRIDDESC files
    used by SMOKE and CMAQ

27
Unidata Integrated Data Viewer (from the IDV
Users Guide)
28
Unidata IDV Globe Display(from IDV Visualization
examples)
29
Unidata Integrated Data Viewer
  • New tool that supports most data requirements
  • Data types include gridded data, Shapefiles,
    radar, satellite, observations
  • Reads remote and local files
  • Understands 3-D, time varying data
  • Currently supports all display requirements
    except possibly polygon-based thematic maps
    (although time series could be improved)
  • Global and flat 2D and 3D displays
  • Animation over time and levels, data probes,
    Crude time series, vertical profiles,
    time-height, HTML viewer
  • 3D displays Topography, isosurfaces
  • Skew-t plots, contours, color filled contours,
    filled grid cells, continuously shaded gridded
    data, wind vectors, wind barbs, streamlines,
    surface met observations

30
Unidata Integrated Data Viewer (ctd)
  • Supports most basic functional requirements
  • animation through time space
  • subsetting, derived parameters, probing
  • zoom, pan, rotate
  • integrated time series other 2D plots
  • May not support many GIS-like functional
    requirements
  • Supports most system requirements
  • Java tool based on VisAD
  • Output as JPEG, quicktime movies, flythroughs
  • Interactive analyses only right now, but
    considering batch mode
  • Some support for running over web / creating web
    pages
  • Available from http//unidata.ucar.edu

31
Air Quality Forecast of Ground Level Ozone 1-hr
average (courtesy NOAA)
OZONE (parts per billion)
32
Spatio-temporal Estimated Data BMELib/Matlab
Visualization(courtesy Dr. Marc Serre, UNC
Chapel Hill)
33
Point Data with PDFs and Computed Mean (courtesy
Dr. Marc Serre, UNC Chapel Hill)
34
Summary
  • Currently available tools do not meet all
    visualization requirements for the UV Tool
  • Unidata IDV comes close
  • Need to consider how we can meet the remaining
    needs
  • Visualization gallery was suggested to share
    ideas among community
  • check CMAS web site for future details
  • for now, e-mail candidate visualizations (lt 3MB)
    to eyth_at_unc.edu)
Write a Comment
User Comments (0)
About PowerShow.com