error when attempting a scatter analysis

[kbedka]

I tried out the nightly today. When I create the scatter analysis, I still see some red squares at the center of the left and middle panels. If I pan to the right, I do see the map with the image data, but there are some strange magenta lines over the top of which I cannot get rid of. See attached image

[tomw]

Kris --

When I use your file:
D:\DATA_DRIVE\kris\NSS.GHRR.NF.D86275.S0857.E1038_NASA_LaRC_cloud_properties_3chan5.nc
and select the "type" of "Grid files (netCDF...."), then select the 11um temperature and "Scatter Analysis" and then select the 3.7um, I get the graph here -- with no zooming/roaming/etc.

(The "magenta lines" you see in your plot are actually the temperature readouts of the point when the little red "cursor" dot is located...but the font size makes it obscure.)

If you just display the 11um temperature data, does the "wire-frame box" appear by itself (as before) but when you zoom out and pan right, you see the data? If so, then you might not have the updated netCDF file....

[kbedka]

this particular file works as you've shown in your graphic. When I try a different file, also over the South Pole, the "wire-frame box" appears by itself (as before) but when I zoom out and pan right, I see the data. See attached graphics. The file I'm referring to is located at the SSEC ftp site:

pub/incoming/NSS.GHRR.NN.D08275.S0459.E0652_NASA_LaRC_cloud_properties_CERESmap.nc

[tomw]

Kris -- I can confirm your observations. I have sent this information to the netCDF-Java developers at Unidata. I will let you know when I hear back from them.

[kbedka1]

My colleague has again run into the issue where she cannot probe a polar orbit cloud retrieval netcdf file due to the issues we've been discussing in this thread and in the following thread (from 2011):

viewtopic.php?f=24&t=781

This is a roadblock to her progress since she needs to know values at individual satellite pixels. Has there been any new developments to rectify this problem? Last I had heard, some work needed to be done by Unidata to fix an issue in NetCDF.

[joleenf]

Hi Kris,

There is a mask function in JPythonMethods that can be turned into a formula in McIDAS-V1.3. This is also already available in the list of formulas in McIDAS-V 1.4. From the jython shell type

Code: Select all

a=selectData()
b=mask(a,'ne',-9,1)

# display the mask in the active window
maskLayer = activeDisplay().createLayer('Image Display', b)

Using 1 ('True') as the last argument returns a field of 1 and NaN.

See mask documentation at http://www.ssec.wisc.edu/visad-docs/jav ... thods.html

In the McV1.4 formula, the mask is applied to a chosen field so that the values of data are returned where the original data is 1. (if I mask any 3.9 micron values greater than 300 K and apply it to 10.7 micron values, the displayed field is are the 10.7 micron values only where the 3.9 micron field is greater than 300 K).

Joleen

[kbedka1]

so would invoking this formula allow my colleague to command-> left click on the image and see the values at the bottom of my display window, similar to an alt-D in McIDAS-X? This is the capability we're looking for with these NetCDF-based AVHRR cloud retrieval swaths, a capability that works with just about all other image-based data.

[joleenf]

Hi Kris,

If I understand what you are asking, yes. Once the data is in the display window, you can use the data probe, so you could achieve that either by displaying via the formula or the jython shell (but for that info, you would display the field*mask from the jython shell). It becomes a style preference.

Joleen

[kbedka1]

I tried your example with the following file

pub/incoming/NSS.GHRR.NF.D86275.S0716.E0901_NASA_LaRC_cloudmask.nc

The data field I selected was Observed 11 Micron Temperature

My display becomes a field filled entirely with 1, as expected, because all Temperatures are greater than -9. When I probe the data with command - left click, I still get nothing except the lat/lon location of the cursor. I'd expect to see a cursor readout of 1 everywhere.

[joleenf]

Hi Kris,

I tested with the same file and noticed the same thing. I then used a file on my machine (unfortunately not the same cloud retrieval algorithm) over Michigan and the probe works. Is this a problem with the probe at the poles?

Joleen