NPP ATMS in Mc-V

[kbedka1]

Can one display Suomi NPP ATMS radiances/brightness temperatures in McIDAS-V?

[bobc]

Hi Kris -

Using both stable 1.6 as well as today's 1.7beta1 nightly, I was able to load in and display some ATMS data ordered from CLASS:

GATMO-SATMS_npp_d20170605_t1908560_e1916556_b29048_c20170606172759645421_noac_ops.h5

This file is loaded through the Under Development > Imagery - Suomi NPP chooser. One of the fields available for display is "ATMS-SDR_All/BrightnessTemperature", but there is no field for Radiance. I inspected this file in ToolsUI and there doesn't appear to be any mention of "radiance" in the file.

For Brightness Temperature, you can use the standard "Image Display" or the "MultiSpectral Display". Through the Layer Controls tab of the Data Explorer, the "MultiSpectral Display" allows you to change the channel/wavelength being plotted in the Main Display.

If you have any questions about this, please let me know.

Thanks -
Bob Carp
McIDAS Help Desk

[kbedka1]

Thank you, could you also inform me of the procedure to view AMSR-E brightness temperature data? I've tried a variety of options, the only data type that seems to work is grid files, but then the data is treated like point data with 2000 time slots at 2 sec resolution. If I select all times, my display takes forever to load and I get a bunch of negative numbers in the display that looks like noise. Any other options? I recall being able to look at AMSR-E in the past and I did not encounter the issue I encounter here

[bobc]

Hi Kris -

Your best bet would be setting up a local ADDE dataset. There are three local servers designed for AMSR-E data:

• AMSR-E L 1b
• AMSR-E L 2a
• AMSR-E Rain Product

I just went and downloaded some AMSR-E L2A Brightness Temperature HDF files from the NSIDC and set up a local server using the "AMSR-E L2a" format. From this, I was able to display brightness temperature data.

https://n5eil01u.ecs.nsidc.org/AMSA/AE_L2A.003/2011.10.04/

If you give this a try and run into any problems, please let me know.

Thanks -
Bob

[kbedka1]

thank you, I tried this and got a very weird display, see attached. Here's an example file on your ftp server,

pub/incoming/AMSR_E_L2A_BrightnessTemperatures_V12_200604061816_D.hdf

If I allow the display to remain in swath LALO projection, it still looks strange. for reference this was the 36.5 Ghz channel. I see similar behavior with other channels

[bobc]

Hi Kris -

I'm replicating your striped display on OS X. I tried the same data on Linux and Windows and the striping problem does not exist there. This striping issue on OS X is referenced in a couple inquiries (1146 and 1014). However, both of these inquiries have a status of "Released" and the problem still exists. I'll look through my notes and see if I can determine why this problem wasn't resolved, and I'll likely write up a new inquiry.

On OS X, you can get your display looking its best by doing the following:

1. Display your data at full resolution
2. To remove the horizontal streaks of data at the bottom of the display, disable fast rendering for the layer. You can do this by right-clicking on the Legend Label of the layer (the blue text in the Legend) and select Edit>Properties. At the bottom of the Settings tab of this Properties window, uncheck "Use Fast Rendering" and click OK. Note that this horizontal streak of data appears on all platforms if fast rendering is enabled.
3. Currently, the only way that I'm aware of to remove the vertical stripes of data on OS X is by adjusting the pixel sampling of the layer. From the Layer Controls, change the Pixel Sampling to an odd value, such as 1, and the striping should go away.

If you give this a try and still run into problems, please let me know. I'll update you if I come up with a workaround to get the data to display at full resolution on OS X or if there are any updates to the inquiry I plan on writing.

Thanks -
Bob

[kbedka1]

I think I've done what you requested but I still get an odd looking display. See attached images. I have my data at full resolution without fast rendering and have an odd number of pixels in the sampling (i.e. 1). Zoomed out you get a psychedelic linear pattern, zoomed in, these are vertical stripes that make the data hard to interpret.

I've also found that on this forum site, I always have notify me when a reply is posted, but I never get notified of your replies. Is my email address correct for my account? kristopher.m.bedka@nasa.gov?

[bobc]

Hi Kris -

Sorry you are still seeing the striping issue, I'm working on replicating it on my end. Would you mind doing some troubleshooting that might help point to the problem? There are a couple things you can do:

1. With your session as-is with the striped displays, can you send me screenshots of your "Formats & Data" and "Advanced" tabs of your User Preferences? There are a variety of settings in there that could possibly impact how the display is created.
2. For clarity, do any of the Pixel Sampling values in the Layer Controls allow the striping to go away?
3. Could you try starting up McIDAS-V in a fresh session? You can do this by using the -userpath flag at startup. For example, from a terminal, run:
   ./runMcV -userpath /Users/kbedka/ATMS_testing
   ...from your McIDAS-V-System directory. This will start up McV as if it were the first time run on your machine. Please re-create your local dataset and try displaying your data again at full resolution with a pixel sampling of 1. Does this help at all?

By the way, Barry is looking into the username/email issue and he should be responding about that shortly.

Thanks -
Bob

[barryr]

Code: Select all

I've also found that on this forum site, I always have notify me when a reply is posted, but I never get notified of your replies. Is my email address correct for my account? k...@nasa.gov?

Hi Kris,
As you Bob mentioned, I checked into the email address issue. It turns out that you have two forum accounts with different email addresses:

• kbedka uses the NASA email address you specified in the post above
• kbedka1 uses your old SSEC email address

You can change your account settings (including email address) by going go the User Control Panel (link in upper-right) then clicking Profile, then clicking Edit Account Settings. So you can do that to check each account's email address and change if desired. However, I don't think you can change both of them to use your NASA email address since it appears that the board only allows a given email address to be linked to one account.

If you have an alternate form of your NASA email address (e.g., john.doe@nasa.gov and johnd@nasa.gov are aliases) you might be able to use those two email addresses in your two forum accounts (kbedka and kbedka1). Alternatively, we could remove the account you don't want/use and leave its posts so they're not lost.

In any case, when you're done you might want to edit the post above and change or remove the NASA email address so it's no longer visible on this webpage.

Barry

[kbedka1]

I attached my screenshots

No the pixel sampling does not cause the vertical striping to go away, no matter what odd number I choose

I tried starting a fresh session, I recreated my local server, displayed the data, pixel sampling of 1, kept in swath projection and also tried reprojecting to "World". Still get the vertical stripes.