<font face="Verdana, Arial, Helvetica" size="2">Michael Bietenholz updated #11645<br />
---------------------------------<br />
<br />
Help with AIPS TYAPL, TYSMO and SY-table values:<br />
------------------------------------------------<br />
<br />
<div style="margin-left: 40px;">Ticket ID: 11645</div>
<div style="margin-left: 40px;">URL: <a href="https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/11645">https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/11645</a></div>
<div style="margin-left: 40px;">Name: Michael Bietenholz</div>
<div style="margin-left: 40px;">Email address: <a href="mailto:mbieten@yorku.ca">mbieten@yorku.ca</a></div>
<div style="margin-left: 40px;">Creator: User</div>
<div style="margin-left: 40px;">Department: AIPS Data Reduction</div>
<div style="margin-left: 40px;">Staff (Owner): -- Unassigned --</div>
<div style="margin-left: 40px;">Type: Issue</div>
<div style="margin-left: 40px;">Status: Open</div>
<div style="margin-left: 40px;">Priority: Default</div>
<div style="margin-left: 40px;">SLA: NRAO E2E</div>
<div style="margin-left: 40px;">Template group: Default</div>
<div style="margin-left: 40px;">Created: 08 February 2018 04:39 PM</div>
<div style="margin-left: 40px;">Updated: 08 February 2018 04:39 PM</div>
<div style="margin-left: 40px;">Reply due: 12 February 2018 04:39 PM (4d 0h 0m)</div>
<div style="margin-left: 40px;">Resolution due: 04 November 2020 12:00 AM (999d 7h 21m)</div>
<br />
<br />
<br />
I'm just wondering about the use of AIPS TYSMO, and EVLA Psum, Pdif and Tsys<br />values (SY-table) which I have little experience with.<br /><br />I'm looking at P-band data 14-DEC-2014 (14B-373). Now I'm observing the Crab, so I expect the source to raise the system temperatures, but that should cause high but not variable Tsys. I loaded the data into AIPS using BDF2AIPS (31DEC18; obit-1.1.500-1-64b). I gather that I should apply gains derived from the SY table using TYAPL, optype 'PGN'. However looking at the Tsys values makes me think I should edit / smooth them first.<br /><br />The TSYS (PSYS) values in the SY table are pretty wildly variable - I attach an example plot from SNPLT: psys-1.png, showing "PSYS" for antennas 1-4, poln=X, all IFs (spw). Antenna 1 has many extremely high values (>1 MegaK) across many IFs and there are sporadic high values in other antennas mostly confined to a few IFs.<br /><br />Zooming in to the above by plotting only values <1000K, I get psys_2.png - concentration of values around 200-600K (except for antenna 1), but quite noisy especially in some IFs (although it doesn't seem to be the RFI-dominated IFs)<br /><br />Zooming in in time, I have psys-3.png, which shows short (~2min) excursions up to ~2000 K for antenna 4,5 in IF16 (472 MHz dark blue).<br /><br />I further attach psum.png - scaled so that sporadic high values up to ~50000 counts are not visible, but more systematic structure is. Also, pdif.png: again, sporadic outliers between ~-1000 and + 1500 counts or off the plot.<br /><br />Finally, there is something very non-random going on in psys-dual.png, psum-dual.png and pdif-dual.png These show only 5mins of data X-pol, IF15. There are periodic, every ~10s, values that are systematically different - even though this is all one source. I see the same pattern over much of run, so its not just over these 5 mins.<br /><br />It loooks like some kind of median filter on fairly short (say 30-sec) timescales is called for. Should I edit out excursions such as seen in psys_3.png? Should I just radically clip SY values antenna 1 (or just throw it out)? Any idea what is going on with the ~10-s periodicity? Any experience with APARM,CPARM values to use for TYSMO? When should I edit out visibility data?
<br />
<HR style="margin-bottom: 6px; height: 1px; BORDER: none; color: #cfcfcf; background-color: #cfcfcf;" />
Staff CP: <a href="https://help.nrao.edu/staff" target="_blank">https://help.nrao.edu/staff</a><br />
</font>