[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #QGM-103347]: Temperature off by a factor of 10 with area files

Hi Jess-

> We noticed an issue with the temperature being off by a factor of 10 in
> 2.7a1 from area files saved by using the "Make Data Source Local" option
> from satellite data.  If you load a satellite image (I used a GOES
> 10.7um) with a "Data Type" of Temperature and make the data source
> local, when you load the area file through the file chooser and display
> the image, the probe lists the temperature incorrectly (see attached
> gif).  When probing the image that is directly from the chooser, the
> temperature is correct.

Wow, this really had me stumped. We are just writing out the bytes
from the ADDE aget into a file.  Nothing had changed on our end for
the writing of the data.  However, we recently added a new 
ImageGridDataSource that would read "images" as grids through
the netCDF-Java Common Data Model.  The file pattern for this new
datasource included .area.  So, that data source is being used to
read in the saved image and obviously has a bug in the reading.
I wasn't accounting for the calibration scaling (10 in this case).
I'm working on a fix for that.

In the mean time, I checked in a new datasource.xml with the .area
removed from the file pattern for the ImageGridDataSource.  That way,
the .area files will be back to using the AREA data source. As a work
around, you can set the Data Type to be McIDAS AREA Files instead of
using I'm Feeling Lucky.

> This was first noticed in our unstable McIDAS-V version which is using
> jar files from 2.7a1 from Feb. 18th.  I tried the nightly IDV 2.7a1 from
> Feb. 23rd and see this problem as well.  I tried 2.6u2 and did not see
> this problem.

Yeah, I changed datasource.xml in the 2.6u2 branch so it wouldn't
pick up AREA files as the new datasource.

Don Murray

Ticket Details
===================
Ticket ID: QGM-103347
Department: Support IDV
Priority: Normal
Status: Open