Constraints - Definitions and Tips
There are two major groups of constraints for Spitzer: Operational Constraints and Observer-Imposed Constraints. These pages give a quick overview of the constraints.
Table Of Contents:
Operational Constraints are those which involve routine telescope operations. They are imposed by the observatory hardware, safe operations requirements, or constraints inherent to the spacecraft and science instruments (including instrument calibrations). These cannot be altered by an observer, unlike the observer-imposed constraints. For more discussion of operational constraints in general (including more figures and some examples), see the Warm Spitzer Observers Manual, especially Chapters 3-5.
Spacecraft constraints: Visibility constraints:
Due to several operational requirements such as solar avoidance and solar panel illumination, Spitzer is constrained at all times to point the telescope no closer than 82.5 degrees toward and no further than 120 degrees away from the Sun. This defines the Operational Pointing Zone (OPZ) of the Observatory. Experienced Spitzer observers should note that the OPZ was reduced by 2.5 degrees in January 2004. This change allowed us to increase the maximum slew rates used by Observatory, which gave us faster slews and therefore more time to integrate on the sky.
The shortest period that any object is visible in the OPZ is approximately 80 days per year (in two intervals of 40 days, separated by 6 months) for targets near the ecliptic. This increases to approximately 120 days per year at +/- 45 degrees ecliptic latitude, to year-round visibility very near the ecliptic poles. This region around the ecliptic poles is referred to as the Constant Viewing Zone (CVZ). For more information, see section 3.2 of the Warm Spitzer Observers Manual.
The roll angle of Spitzer is at all times precisely determined by the spacecraft pointing constraints. The sunshade on Spitzer must always be kept within 2 degrees of the radial direction to the sun. Because of this, the orientation of the focal plane on the sky is a function of calendar date and the ecliptic latitude of the object. This orientation changes by approximately 0.5 to 1.0 degrees per day, depending upon the ecliptic latitude of the target. The result is that if spectra of an object near the ecliptic poles were taken using the same slit separated by one week, then the orientation of the slit on the object would differ by 7 degrees. In the case of point sources this should not be a concern, but in the case of extended sources (such as nearby galaxies), observations which are not taken in sequence will sample different regions of the object.
Instrument and spacecraft constraints include:
- Star tracker alignments must be performed every 8 hours.
- Downlinks (communication between Spitzer and Earth) occur approximately every 24 hours. The length of downlinks can vary from 90 minutes to several hours depending on the Deep Space Network (DSN) schedule.
- Inertial reference unit (IRU) calibrations must be performed every few weeks, and star tracker to PCRS calibrations are done about once every 12 hrs (no more than about 16 hrs apart), usually in conjunction with a downlink.
- IRAC Post-Cryo AORs are restricted in length to 24 hours. This limitation is automatically imposed by Spot. However, IRAC PC observations in general should not exceed ~ 12 hours due to the necessary star tracker calibrations (see above). The 24 hour limit is to allow the use of AORs for extended duration science observations, such as exoplanets, without the SSC having to convert them into Instrument Engineering Requests (IERs), as was necessary with the old 8 hour duration limit.
Spitzer allows observers to request constraints on their observations. Because observer-imposed constraints (combined with the operational constraints) make it more difficult to schedule observations in an efficient manner, it is essential that observers keep constraints to a minimum. It is recognized that some scientific programs can only be accomplished through use of observer-imposed constraints; however, these must be thoroughly and soundly justified in the observing proposal for consideration by the proposal review panel. We refer observers to the Spitzer current Call for Proposals for further discussion of observer-imposed (also called "user-imposed") constraints. For more discussion of observer-imposed constraints (including how to apply them), see the Warm Spitzer Observers Manual, specifically Chapter 5, or the Spot User's Guide, Chapter 13.
The constraints available to observers are:
- Chain (ordered, non-interruptible group)
- The AORs will be executed in the order specified with NO interruptions in the chain. If you chain together AORs, you must ensure that the chain does not exceed the maximum allowable duration for a single AOR (6 hours for IRAC). Spot will warn you if your chain is too long. It will show you the total time in the chain in the constraints window.
- Sequence (ordered, interruptible group)
- A sequence constraint is similar to, but less stringent than, a chain constraint. The AORs will be executed in the order specified and a duration in which they should be completed is specified. The sequence constraint should only be used when the science requires sequential ordering of the AORs. For AORs in which the order of observation is not important, a "group within" constraint (see below) should be used instead of a sequence constraint.
- A group-within constraint specifies that a group of AORs will be executed within a specific length of time but with no particular starting date/time constraint. Once the first AOR has been executed, the rest of the AORs in the group will begin within the specified time interval. They may be executed in any order within the time interval. This is similar to a sequence constraint, but the observations may be executed in any order.
- Timing constraints consist of defining a window or series of windows for the start time of an AOR. If the open and close times of the window are specified to be identical for a moving target, then the AOR will be scheduled as an absolute time observation at that time, and will be executed at that time or no more than 3 seconds later. Spitzer's scheduling architecture generally operates on relative time, so for inertial targets, the (inertial target) AORs will simply run in order. Timing constraints for inertial target AORs should be macroscopic (days, weeks, months), not microscopic (seconds, minutes, hours). Spot will warn you if you set a timing constraint such that the target is not visible for some portion of time within the timing window, but will still allow you to set the constraint.
- A follow-on constraint executes the 'follow-on' AOR within a specified time range after a particular initial AOR has been executed. (In other words, from the end of the first AOR to the beginning of the second AOR.) An instance in which this constraint might be used is when the observer wishes to perform an IRS "peak-up only" observation to confirm target placement before starting (perhaps much later) a second observation to actually obtain spectra of the target. This constraint could also be used for periodic observations of a target where the interval between observations is relatively short (hours to a small number of days).
- The shadow constraint is a special case of the follow-on constraint, and is used to obtain background measurements for moving targets. The primary AOR is executed as specified. The shadow AOR will be executed to repeat the track of the primary observations. The selected AOR parameters must be identical in the two AORs. The shadow may be executed before or after the primary AOR. Note that the shadow does not re-observe the target at a later date, but rather the background of the primary observation.
Observer-Imposed Constraints: Tips and Tricks
Observer-imposed constraints can severely limit the scheduling efficiency of the telescope. With that in mind, observers should note that all constraints need to be strongly justified in the telescope proposal. With scheduling efficiency in mind, the following constraint tips are offered. For additional insight into scheduling, see this overview of how scheduling works.
Make constraints flexible!
- In general, "GROUP-WITHIN" is more likely to be scheduled than a "SEQUENCE" which is more likely to be scheduled than a "CHAIN."
- A "GROUP-WITHIN" that constrains 4 hours of observations to occur in the same week is more likely to be scheduled than a "GROUP-WITHIN" that constrains the same 4 hours of observations to occur within a 12 hour period. A good rule is to have the window of the constraint be at least twice the total duration of the AORs within the constraint.
- A less tightly mandated constraint allows the Scheduling Team more leeway to solve problems and avoid conflicts, as well as produce more efficient schedules. For example, do not use a "SEQUENCE" if a "GROUP-WITHIN" will do.
- Extremely complex constraint structures formulated to save a small amount of observing time may render the observatory LESS efficient rather than more so. Please let our schedulers take care of optimization.
- Narrow timing constraints (less than 1-2 days) are more difficult to schedule than timing constraints with windows of ~ 10 days. And timing constraints with less than a 1 day window can be challenging to impossible to schedule. Use these constraints with care.
- When using shadow constraints, it is easier to schedule observations if your shadow (background) observation comes AFTER the main observation. If it doesn't matter to your science whether your shadow observation comes before or after the object, put it after.
Use simple constraints!
- Constraints may be combined, but do so judiciously.
- The more complex a set of constraints, the lower the probability that it can be scheduled.
Be aware of the operational constraints!
- Spacecraft and instrument-imposed events take place roughly every 12 hours. AOR durations should, in general, not exceed this unless there is a strong science justification (i.e. a long duration exoplanet observation).