Background:
What I am really trying to do is to run the *CrIMSS EDR* with on-orbit data, but in order to do that I need a terrain-corrected CrIS SDR geolocation file. Since these are not delivered from the IDPS, that puts me into running the CrIS SDR to try to generate the one that I need. In trying to get the CrIS SDR running with real data, I had quite a few difficulties with the TLE file. I will not recount the story of my troubles; rather, I will just record here what I learned.
What I learned:
1. The TLE file is *not* an optional input to the CrIS SDR, even if you have the spacecraft diary. This is already documented, for example, in the CrIS SDR OAD. Moreover, the config file ProSdrCris_CFG.xml does not list the TLE file as optional, but you might not believe because it is counter-intuitive. Believe it.
2. When you use the ADL Unpacker to unpack the HDF5 version of the TLE file, you will not get an ASCII metadata file. This appears daunting but is not actually a big deal. You can use the sample TLE ASCII metadata file that comes with the ADL distribution. Just edit it so that the filename displayed in the metadata file is the same one that you unpacked. I do not think that the path matters, only the filename. You can leave the URID in the metadata filename alone.
3.The unpacked TLE file will have a name like, for example,
TLE-AUX_npp_20120309154910Z_20120309000026Z_ee00000000000000Z_-_c3s-_ops_all-_ops.msd
For your input, you should prepend the URID to this filename, as
4e89df64-5b242-9b9dea83-ea2d2a09.TLE-AUX_npp_20120309154910Z_20120309000026Z_ee00000000000000Z_-_c3s-_ops_all-_ops.txt_Internal
Notice that I also changed the filename extension on the unpacked file from "msd" to "txt_internal".
4. Finally, there is a problem of file size. The code is expecting the TLE file to be 142 bytes long, and that is the size of the sample file that is provided with the distribution. However, when you use the Unpacker, the TLE file size that you get is 140 bytes. The result is that the code fails. To remedy this problem, you have to add two blanks before the final newline. I do not know what the source of this discrepancy is. [NOTE ADDED MARCH 23: The original post was on March 21. Bryan is right in his comment (see below). This was indeed covered in the workshop but I did not remember it. What I wrote above also is slightly incorrect. You need to add one blank to each line, not two blanks to the second line. The slide that Bryan refers to has it right. Anyway, it doesn't hurt to have the information in more than one place.]
TLE file in the CrIS SDR: What I learned the hard way
TLE file in the CrIS SDR: What I learned the hard way
Last edited by rcember on Fri Mar 23, 2012 10:38 am, edited 1 time in total.
Re: TLE file in the CrIS SDR: What I learned the hard way
Richard, for item 4:
4. Finally, there is a problem of file size. The code is expecting the TLE file to be 142 bytes long, and that is the size of the sample file that is provided with the distribution. However, when you use the Unpacker, the TLE file size that you get is 140 bytes. The result is that the code fails. To remedy this problem, you have to add two blanks before the final newline. I do not know what the source of this discrepancy is.
This discrepancy in the size is an artifact of the ING MSD process not being available to process the file for ADL. ING MSD will be available in ADL with the ADL 4.0 release, but in the meantime as Richard mentions one needs to add the two spaces before the final new line. What ING is actually doing is null terminating the strings and I was told padding out the structure to be 4 byte aligned. Even though a minute conversion, this is a conversion that occurs in the ING MSD software operationally.
We covered this in the ADL 3.1 Workshop in January in slide presentation, though I'm sure one could have easily missed this:
07_ADL_Workshop_Running_CrIMSS_Chain.pptx or 07_ADL_Workshop_Running_CrIMSS_Chain.pdf on slide 33.
Bryan Henderson
Raytheon Company
4. Finally, there is a problem of file size. The code is expecting the TLE file to be 142 bytes long, and that is the size of the sample file that is provided with the distribution. However, when you use the Unpacker, the TLE file size that you get is 140 bytes. The result is that the code fails. To remedy this problem, you have to add two blanks before the final newline. I do not know what the source of this discrepancy is.
This discrepancy in the size is an artifact of the ING MSD process not being available to process the file for ADL. ING MSD will be available in ADL with the ADL 4.0 release, but in the meantime as Richard mentions one needs to add the two spaces before the final new line. What ING is actually doing is null terminating the strings and I was told padding out the structure to be 4 byte aligned. Even though a minute conversion, this is a conversion that occurs in the ING MSD software operationally.
We covered this in the ADL 3.1 Workshop in January in slide presentation, though I'm sure one could have easily missed this:
07_ADL_Workshop_Running_CrIMSS_Chain.pptx or 07_ADL_Workshop_Running_CrIMSS_Chain.pdf on slide 33.
Bryan Henderson
Raytheon Company