Computer Assignment #4
Assigned: 09 Feb 2004
Due: 16 Feb 2004
For this assignment, we will be investigating the differences in hand analysis and computer generated analysis. We will be using two GEMPAK programs: sfcntr and gdcntr.
One of the issues regarding the analysis of real data is that the data is irregularly spaced. Most graphing routines like to have data in a regularly spaced grid. One method of for making the transition is to use the Barnes Objective Analysis. Take a look at the link and read through what GEMPAK does and what are the meanings to gamma and npass.
New parameters for sfcntr.
CINT |
LINE |
MAP |
NPASS |
CNTRPRM |
data used is the standard surface data file
New parameters for gdcntr
| CINT |
CONTUR |
LINE |
MAP |
CTYPE |
GFUNC |
data used is the surface grided data
Tasks
- Analyze the map handed out in class for pressure using 4 mb intervals. Use 1000 mb as your base level contour. Be honest and do this portion of the assignment first.
- Plot a surface analysis map for 12Z 26-Jan-2004 using the Gempak routine sfcntr.
Contour the mean sea level pressure for using an interval of 4 mb. Experiment with various values of npass and filter.
- Now plot the same analysis using the Gempak routine gdcntr.
Questions:
- Briefly discuss the advantages and/or disadvantages of using computer generated analysis.
- Is there an advantage of using more passes of the Barnes scheme?
- Would turning off the filter help improve the analysis?
- What scales would be better served by a hand or computer generated analysis?
- What, if any, difference is there in using sfcntr and gdcntr?
- What, if anything, do you know about the data sets you are using?
- Does one program have more advantages than the other?
- Does one program cover more cases than the other?
- What is missing from the plot when using gdcntr? What procedure (program) would one need to implement in order to make comparable plot to using sfcntr?
Hints:
- Grab the file /home/metr2413/conus.nts and put it in your personal directory for setting the map area and map projection to make for a nice map background for the contiguous United States.
- Using a line type of dash or dot for the map background can make the resulting plot easier to read.
- Another way to help un clutter the analysis is to pick a color equal to the background so that the data will not be viewed. For xw, color = 32 (black) makes it easier to see on the screen while color = 31 (white) will allow for a lighter gray when using ps.
- Make as many plots as you need to support your discussion.
- It might help to overlay plots (clear = no) and vary the line type to aide in the comparisons.