Preliminary testing of ADL CrIS SDR 3.0

Issues related to the CrIS SDR algorithm and data
Post Reply
tremblay
Posts: 1
Joined: Wed Aug 17, 2011 12:00 pm

Preliminary testing of ADL CrIS SDR 3.0

Post by tremblay »

Greetings,

The ADL CrIS SDR 3.0 test case ran successfully. After creating plots,
several values appears to be incorrect. They are:

1) Earth scene ZPD magnitude for LW and MW are zero (most, not all).
2) One SDR FCE count in LW is -1, this value is suspect because the simulated data did not included any FCE.

The provided test case has 9 input RDR files in DMS format (or IDPS save set format) each containing 1 granule
composed of 4 scans. Only the middle granule is processed by the executable.

Hope this helps.

Denis Tremblay
bhenders
Posts: 72
Joined: Wed Jan 05, 2011 9:27 am
Location: Omaha, NE

Re: Preliminary testing of ADL CrIS SDR 3.0

Post by bhenders »

Denis,

Can you clarify that the results issues are from your ADL run or whether they also exist in the dropped truth results. I'm guessing the latter and a quick look at the FCE count in the dropped results show the "-1". I'm guessing your assertions are not ADL issues but the validity of the IDPS output.

9 input RDR files were provided because you need 4 cross granules (15 scans) on either side to fill the sliding window for CrIS in order to process the one granle.

Thanks,

Bryan Henderson
Raytheon Company
bhenders
Posts: 72
Joined: Wed Jan 05, 2011 9:27 am
Location: Omaha, NE

Re: Preliminary testing of ADL CrIS SDR 3.0

Post by bhenders »

Denis,

We have looked at the truth CrIS SDR granule that was dropped with ADL 3.0 and the ZPD magnitude values in that truth granule have no zeros in the MW and LW bands. There are two zero values in the SW band. The min and max for this field for all bands range from 0 to 65520. Please note that this field is an unsigned integer.

The SDR FCE count field is also an unsigned integer field and the (-1) that you are seeing is actually a 65535 value which is denoted as a not applicable fill value for the field, which you can see in the XML legend for the field. According to the data quality team that value is appropriate for this granule for the truth results we were delivered.

Bryan Henderson
Raytheon Company
Post Reply