Non-Sidereal Guiding Instructions for LBT

Details for sending a hijacked Gregorian Non-Sidereal preset in GUIDE or ACTIVE mode

This section assumes that you want to use off-axis guiding on a non-sidereal target so that slit spectroscopy is practical. This is generally used for non-sidereal objects with drift rates between 1 and 100 arcsec/hour. (The best to date is when we observed Quasi with MODS in 2017 with rates of 140 arcsec/hour, but that required a manual modification of the wfs exposure times (editting config files in advance). See IT 6312 for discussion of the wfs exposure times.) These instructions describe how to generate a catalog with current guide stars (guide stars appropriate for the current position of the target) and how to send the preset. WHILE WE OBSERVED QUASI IN SPRING 2017, THIS FUNCTIONALITY WAS BROKEN IN THE 2017A BUILD. A PATCH TO FIX IT 7081 was installed on 20180307 UT. See more recent IT 7643.

Preparation

  • Create an LBT/JPL Ephemeris File for this target for the particular UT night with NSIGUI (part of the LBT TCS). (The NSIGUI must be run from a mountain workstation. NSIGUI queries go through the Horizons telnet interface.)

  • Create a LBT Catalog File (used by IRAF.LBTtools.Observe.newpoint and IDL presets) with guidestars at the appropriate time of the observation with /home/lbcobs/supportscripts/NonsiderealSupport/nscatalog.py
  • This example selects guide stars for asteroid Sappho at 02:00 UT on 20141020 UT.
    • to create: /lbt/catalogs/Scratch/catalogs/Sappho_20141020_0200.catalog
    • Usage: /home/lbcobs/supportscripts/NonsiderealSupport/nscatalog.py -d 20141020 -e 2:00 -t Sappho

  • (Alternative for LUCI, Al Conrad has a script that generates multiple Scriptor XML files for a moving target.)

Sending the Preset

  • Choose File on the NSIGUI to load your pre-existing Ephemeris File (be sure to get the one with the correct UT date)

  • Set Target on the NSGUI to hijack the next preset for the non-sidereal target. (Al reminds us: Hijack the truck before leaving the garage.) "Hijack" in this context means that PCS should use the selected non-sidereal ephemeris as the target rather than the fixed RA/DEC of the target specified in the original preset command. (Note that mulitple NSIGUIs may be open at once, when you preset the Set button, the ephemeris file selected on that NSIGUI where you pressed the button will be loaded into IIF for the hijack. The loaded file will be displayed on all of the NSIGUIs. Thus, you could have a separate NSIGUI open for each of the Galilean moons in order to rapidly hijack presets between them.)

  • Send an ACTIVE preset from IRAF (or IDL) using the Target Name of the catalog file (Sappho_20141020_0200) and a guide star selected from that file. This preset provides the guide star information and the hijack from the NSIGUI provides the NS target information to TCS.

Searching for main belt asteroids that are observable

  • Copy/paste the output of the web-based query to a file.

  • cut -c 1-7 file > filename
to output just the first column which contains the IAU ID number... A minority of the output objects don't have IAU ID numbers. The IAU number search fails for asteroids with number < 11. For the purposes here, the list may include a combination of IAU ID numbers and names, so "Ceres, Pallas, Juno, Hebe" work fine. Alternately use cut -c 8-28 file > filename to get names rather than IAU ID numbers. Names have problems where there are ambiguous names (i.e. Jupiter or Io). See ftp://ssd.jpl.nasa.gov/pub/ssd/Horizons_doc.pdf for other query naming tricks like ASTNAM=Iris .

  • /home/lbcobs/supportscripts/NonsiderealSupport/fastmover.pl filename 2015-Mar-17 7
    • argument 1 is a filename with one potential target per line.
    • argument 2 is the date in the specific yyyy-mmm-dd format (it will default to today if not specified).
    • argument 3 is the integer UT hour for calculating the observability (it will default to UT=7 if not specified).

Warning this outputs a file called ID*eph and ID*out for each object (ID) in the input list. These are put in a subdirectory of /tmp. The final output is "filename.sort" which is sorted by rate (sqrt(rarate**2 + decrate**2) in descending order - fastest to slowest. Only objects above a certain elevation are printed. The list of main belt asteroids that this produces is fine for most purposes, but a flaw in this search procedure is that sbwobs.cgi gives you N asteroids first and does not give the more recently discovered earth-crossing objects with much higher rates. See the similar instructions below for selecting much faster moving objects. Note that there is a pre-made list of earth-crossing asteroids.



Details for sending the LBC Prime Focus Non-Sidereal preset in TRACK mode

This section assumes that you want to image a non-sidereal target with LBC. LBC selects its own sidereal guide stars from what it finds in the field of the technical chip, so no special catalog is needed.
  • LBC guiding is practical for non-sidereal objects with drift rates less than 80 arcsec/hour. This limit depends on the exposure time due to sextractor issues described in IT 7847. For rates of 70 arcsec/hour the exposure time limit was about 140 sec.
  • Even after IT 7847 is fixed, you still have the limit of your guidestars drifting off the tech chip within the exposure time (about 300 arcsec/hour for a 6 minute exposure). You can maximize this limit by selecting a position angle where the sidereal guidestars drift along the long axis of the tech chip.
  • To image non-sidereal targets with faster rates, LBC guiding should be turned off (unchecked on the LBC UI) since it can't keep up.

The LBC OB files presently have no mechanism to send a non-sidereal target, so you must use this hijack method for tracking NS targets with LBC. In open-loop tracking (without guiding) LBC has observed non-sidereal targets moving at up to 1000 arcsec/hr, and much higher rates are possible (see below).

Preparation

  • Create a LBT/JPL Ephemeris File for that target for the particular UT night with NSIGUI (part of the LBT TCS). The NSIGUI must be run on a mountain workstation. NSIGUI queries go through the Horizons telnet interface. The resulting files are stored in /lbt/data/ephemeris/ .
  • Prepare a normal LBC OB which specifies the object name, filters, dither offsets, exposure times, etc. that you want for the non-sidereal target observations. The RA/DEC coordinates may be fictitious (but not DEC=-90), although there is some use to having them near the approximate target coordinates.

Sending the Telescope Preset

  • Click Choose File on the NSIGUI to load your pre-existing Ephemeris File (be sure to get the one with the correct UT date) as the next target.

  • Click Set Target on the NSIGUI to hijack the next preset with the chosen ephemeris file. (Al reminds us: Hijack the truck before leaving the garage. = Click Set prior to sending the Telescope Preset.)

  • Play an OB (with any valid RA/DEC) from the LBC UI to send the telescope TRACK preset. (The FITS headers are nicer if you use the approximate RA/DEC of the target.)



Details of Adaptive Optics Observations of Slow-Moving Non-Sidereal Targets (self-referenced, with dithering) using sidereal presets

For this definition, "slow-moving" means objects with apparent motion slower than about 40 arcsec/hour which is the approximate limit of observations with the hexapods offloading the reflex motion from the thin shell.

For this example, we assume that the target asteroid is also the AO reference object and that we are using it on-axis.

For LBTI or ARGOS, send the normal ACE_TRACK preset to the current coordinates of the NS target (for the target and reference) in sidereal mode. This works fine for things like Io at ~10-15 arcsec/hour, but can cause problems if the drift rate that has to be offloaded is larger than ~40 arcsec/hour.

For LUCI AO, you can send the normal ACE_ACTIVE preset to the current coordinates of the NS target (for the target and reference) in sidereal mode with an off-axis sideral guidestar. (Al Conrad has some scripts to help prepare LUCI XML scripts for the moving target and off-axis guide stars.) We have tested this briefly on 16 Psyche moving at 44 arcsec/hour.

After this observation has completed, you will need to follow the suggestion of Issue 5808 in order to restore normal pointing.
  • The procedure to recover from this condition is to Initialize PSF LUTs on 4 mirrors (primaries and secondaries), ptreset both sides, and to send a new sidereal preset.

What if my AO target is moving faster than ~50 arcsec/hour?

See the next section for slightly faster non-sidereal targets with non-sidereal tracking rates, but still using sidereal off-axis guide stars.

Details of Adaptive Optics Observations of Non-Sidereal Targets (self-referenced, with dithering) using non-sidereal presets with sidereal guide stars

MISSING TEXT HERE

Barry and Enrico used this mode for AO observations of Eugenia in Fall 2020 (28 arcsec/hr on 20201002 UT).
Fri Oct  2 08:29:24.733 2020 59124.35375849262 5 IIF  iif.command.preset.started left  Preset command started for LUCIFER mode ADAPTIVEACE_ACTIVE with guidestars, offset, non-sidereal target, non-sidereal AO reference
2020-10-02T08:29:34.468646+00:00 tcs2 LBT_PCS: setDifferentialRateTarget. modifyTargetPO 1 tai 59124.35386924769409 left 02:58:25.222 +07:59:43.39 RARateCosDec: -0.00212166551335 DecRate: -0.00140622842062 Mode:POSITION Angle:0.00000000000000 Wave: 1.50000000000000 rotIndex: 6

What if my AO target is moving faster than ~120 arcsec/hour?

See the next section for fast moving non-sidereal targets........which must be observed at non-sidereal tracking rates without off-axis guide stars.



- - - -

Observations of Very Fast Non-Sidereal Targets

Self-referenced Adaptive Optics Observations of Non-Sidereal Targets

As of September 2017 (TCS Build 2017A) the PCS is able to accept a non-sidereal target as both the Target and the AO Reference Star. Not all instruments are ready for sending this type of preset, but you are still able to hijack sidereal presets with the NSIGUI. This allows the AO system to lock on a bright-enough-for-AO target using an ACE_TRACK preset at essentially any NS rate. This was well-tested on 3200 Phaethon on 20171215 UT with NS rates of -1700,-700 arcsec/hour. That observation was using hijacked ACE_TRACK presets with LBTI/LBTI.

What did not work on 3200 Phaethon was updating the NS target from Ephemeris file - the average position was OK, but the telescope polynomials had ~1 arcsec jumps. These jumps caused the AO loops to safe skip. See IT 6746 which was fixed with a patch in Jan 2018.

Searching for very fast moving non-sidereal objects

  • Create a text file with names or IAU ID numbers of potential targets (one per line). Note that there is a pre-made list of earth-crossing asteroids mentioned below.
    • entries in input list can be either ID numbers or names: e.g. '291' or 'Alice' or 'C/2014 Q2,' but NOT '291 Alice'
    • note that your search may fail because your target has multiple entries in Horizons - use NSIGUI "View Log" to check for such cases.

  • /home/lbcobs/supportscripts/NonsiderealSupport/fastmover.pl filename 2015-Mar-17 7
    • argument 1 is a filename with one potential target per line.
    • argument 2 is the date in the specific yyyy-mmm-dd format (it will default to today if not specified).
    • argument 3 is the integer UT hour for calculating the observability (it will default to UT=7 if not specified).

Warning fastmover.pl outputs a file called ID*eph and ID*out for each object (ID) in the input list. These are put in a subdirectory of /tmp/. The final output in your local directory is "filename.sort" which is sorted by rate (sqrt(rarate**2 + decrate**2) in descending order - fastest to slowest. Only objects above a certain elevation (20 deg) are printed.

See /home/lbcobs/supportscripts/NonsiderealSupport/earthcrossing.lst for an pre-made search list of ~125 asteroids.

How fast can we go?

How fast can we go? The telescope mount hardware routinely tracks at 15 deg/hour (we know that from stars), so we can track most NS objects moving at 100000 arcsec per hour relative to the stars (with the usual restrictions on zenith crossings from high azimuth velocities).

We believe the limits are:

a) accuracy of the orbit and ephemeris (depends on the object, but usually very good)

b) accuracy of the TCS ephemeris calculation and/or telescope pointing - generally 5-10 arcsec

c) accuracy of the local parameters (the position of LBT is known to a couple meters)

d) ability to follow the rapidly changing ephemeris. As of January 2018 (a patch to TCS Build 2017A) the PCS is able to adjust the rates to follow a rapidly changing ephemeris. (The January 2018 patch to 2017A fixed bugs in IT 6743, 6746). This feature can be toggled on/off on the PCSGUI. Default ephemerides generated by Horizons queries from the NSIGUI have 1-minute granularity, and the pcs.conf file is set by default (2017 thru Jan2020) to update at 0.5 second intervals by interpolating from the ephemeris file.

If you are also trying to do adaptive optics, then there are additional considerations of:

e) how fast the upper atmosphere is whizzing by due to the rapid tracking rate

f) vibration from the rotating enclosure at high azimuth velocities (still a function of zenith distance). This may start to degrade AO performance at azimuth rates above 0.5 deg/sec.

On 20190505 UT we hijacked an LBC TRACK preset to geostationary satellite ANIK, so we were tracking with full anti-sidereal corrections (-54000 arcsec/hr = telescope stopped). The LBC image was nice and round (seeing-limited).

On 20200127 UT we used an ACE_TRACK preset to observe geostationary satellite ISAT901 with the SX LUCI AO system. The non-Horizons ephemeris ran into drift problems and hexapod limits after about an hour.

What if I want a different duration of the ephemeris file?

This command which underlies NSIGUI makes a 24-hour ephemeris file rather that the default 0-15 UT. TCS is expecting 1 minute intervals, but you can query for 1h or 1d intervals.
nsQueryEngine.tcl "Anik" 2020-01-26 00:00 2020-01-26 23:59 1m Anik_2020-Jan-26.txt > Anik_2020-Jan-26.log

What if my target is not in JPL Horizons?

Be sure to read the .log file created during your query. It often contains information about why your target name didn't result in a unique match.

For some geostationary satellites, we don't find them in JPL Horizons. In that case it is necessary to get TLEs (Two-Line Elements or Three-Line Elements) and feed them into Horizons to make the ephemeris. Search for your satellite on celestrak.org and then enter the catalog number into this script.
/lbt/tcs/current/tlQueryEngine.tcl 49817 2024-01-21 00:00 2024-01-21 15:00 1m /lbt/ephemeris/STPSAT-6_2024-Jan-21.txt > /lbt/ephemeris/STPSAT-6_2024-Jan-21.log
See IT 9036. Effective 24-Jan-2024, NSIGUI has a button which collects TLE elements from celestrak.org and then feeds them to JPL Horizons.

-- %USERSIG{JohnHill - 2014-10-22}%
Topic revision: r48 - 12 Feb 2024, JohnHill
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback