SCD User Briefing NCL and PyNGL Visualization for the Geosciences

1
SCD User Briefing NCL and PyNGLVisualization
for the Geosciences
85th AMS Annual Meeting January 9-14, 2005

• Don Middleton with presentation material
  developed by Luca Cinquini, Mary Haley, Fred
  Clare - and many project collaborators
• NCAR Scientific Computing Division
• Section Head, Visualization Enabling
  Technologies
• January 11, 2005

2
NCL and PyNGLVisualization for the Geosciences
3
Topics

• NCAR Command Language (NCL)
• GSUN (NCL visualization interface)
• PyNGL
• Example codes and visualizations
• Future plans

4
What is NCL?

• NCAR Command Language
• Scripting language, powerful in 3 areas
• File input/output
• Data processing
• Publication-quality visualizations
• Runs on Linux, MacOSX, Solaris, IRIX, AIX, OSF1,
  Cygwin

5
NCL File I/O

• Based on the netCDF model
• Read/write netCDF, HDF4, ASCII, binary
• Reads GRIB, including ECMWF complex coefficients
• One interface for several data formats

6
NCL Data Processing

• Over 600 functions and procedures
• Spherical harmonic based calculations
• Scalar and vector regridding
• Vertical interpolation, including hybrid
  coordinates to pressure level
• Spectral/cross spectral and wavelet analysis
• Can access Fortran and C routines directly from
  NCL

7
NCL Visualization

• Publication-quality visualizations
• Contour, line, vector, streamline, map
• Customization hundreds of options
• Output to X11, PostScript, PDF
• Lots of examples

8
GSUN interface to NCL visualization

• Getting Started Using NCL
• Successful collaboration w/CCSM users
• User-friendly, customized visualization functions
• Extensive documentation and examples
• Interface of choice for NCL visualization
• GSUN used as a model for PyNGL

9
What is PyNGL?

• Python NCL Graphics Library
• Python version of popular GSUN interface
• Same publication-quality graphics as NCL
• Utilizes existing Python modules
• Multiple output formats
• Extensive documentation

10
PyNGL
GSUN
import Ngl from Scientific.IO.NetCDF import
Open the NetCDF file. nfile NetCDFFile("meccatem
p.cdf","r") Get lat/lon/temperature
variables. lat nfile.variables"lat" lon
nfile.variables"lon" T
nfile.variables"t"0,, Open a PS
workstation. wks Ngl.open_wks("ps","mecca")
Set contour scalar field resources. res
Resources() res.sfXArray
lon res.sfYArray lat Tickmarks
displayed by default. Size maximized by
default. Draw contours over a map contour
Ngl.contour_map(wks,T,res) Ngl.end()
load "gsn_code.ncl" begin Open the NetCDF
file. nfile addfile("meccatemp.cdf","r")
Get lat/lon/temperature variables. lat
nfile-gtlat lon nfile-gtlon T
nfile-gtt(0,,) Open a PS workstation. wks
gsn_open_wks("ps","mecca") Set contour
scalar field resources. res
True res_at_sfXArray lon
res_at_sfYArray lat
res_at_pmTickMarkDisplayMode Always
res_at_gsnMaximize True Draw contours
over a map. contour gsn_contour_map(wks,T,res)
end
11
Data from the MECCA Project
12
From a presentation of Jeff Kiehl (NCAR) and Tim
Schneider (NOAA)
13
Wind velocity vectors
Data from the MECCA Project
14
Data from the MECCA Project
15
Courtesy Sylvia Murphy, NCAR CGD
16
From John Ertl, FNMOC
17
Based on a visualization of Joel Norris (Scripps)
using synthetic data
18
Climate division data courtesy Mark Stevens, NCAR
CGD
19
Courtesy Mark Stevens, NCAR CGD
20
Contouring on arbitrary triangular meshes
Image from Jonathan Shewchuk, author of Triangle
21
Grid from Tom Gross, NOAA/NOS/CSDL/MMAP
22
Data and grid from Tom Gross, NOAA/NOS/CSDL/MMAP
23
Future PyNGL plans

• Pythonize NCLs powerful I/O module and climate
  and interpolation functions
• Add more functionality and examples

24
PyNGL/NCL Documentation http//www.pyngl.ucar.edu
/ http//ngwww.ucar.edu/ncl/ PyNGL/NCL talk
email listshttp//www.pyngl.ucar.edu/pyngl-talk/
http//ngwww.ucar.edu/ncl/ncl-talk/ Download
PyNGL/NCL http//www.pyngl.ucar.edu/download.html
http//ngwww.ucar.edu/ncl/download.html
25
END