Graphics

Graphics cindyhg Thu, 04/25/2019 - 12:45

The plotting capabilities of MET are intentionally pretty limited. The only real plotting included is the PostScript output of the Wavelet-Stat tool and the PostScript and PNG output of the MODE tool. MET is intended to be a set of command line tools for evaluating forecast quality. Therefore, the development effort is focused on providing the latest, state of the art verification approaches, rather than on providing nice plotting features. However, good plotting tools are needed for visualizing the MET output. The ASCII output statistics of MET may be plotted with a wide variety of plotting packages. The list of plotting tools below are those that we know have been used to plot MET output to date, but is certainly not exhaustive:

  • The R statistical package is a free package that runs on most operating systems and provides nice plotting features and a wide array of powerful statistical analysis tools.
  • The NCAR Command Language (NCL) is a free package that provides nice plotting and analysis tools.
  • IDL is a widely used (but not free) plotting and analysis tool.

Rather than trying to guess user's wishes and develop plotting scripts, we'd like to solicit assistance from the MET user community. If you write a script to make a nice plot of some MET output, we'd greatly appreciate your contributing it back to us. We're actively gathering such example scripts and will be posting them on the sample scripts page. We do request that you provide sufficient comments in your scripts so that other users will be able to adapt them to their needs. Scripts using any of the tools listed above, or any others as well, would be greatly appreciated.

Please send any scripts you'd like to contribute via met_help@ucar.edu.

METViewer

A tool called METViewer has been developed and is used internally by the DTC for plotting MET output verification statistics. The tool utilizes database software and R to generate several different types of plots including series plots with confidence intervals, box plots, x-y scatter plots and histograms. METViewer also includes a web interface that runs on apache tomcat. Limited distribution of METViewer is planned for the future.

The current METViewer web interface for the data processed by the DTC can be found here:

Congratulations!

You have now completed the MET Online Tutorial for version 8.0!

Should any questions arise that are not covered in this tutorial, the MET Users Guide, or the known issues page, please contact us via met_help@ucar.edu. We also appreciate comments and suggestions about the MET package and requests for additional capabilities. Feedback from the community is used to plan future releases, subject to time and budget constraints.

Thanks for taking the time to work through this tutorial!