9.6.1 Constraints

Since each of the $1^{\circ} \times 1^{\circ}$ maps take about three hours, and the field of view rotates in time, we will need to constrain these AORs to occur as close together in time as possible. For example, we do not want the situation that is depicted in Fig. 9.15; in this case there is an irregularly-shaped wedge of missing wavelength coverage between the maps because one was taken at the beginning of the visibility window, and the other was taken at the end of the window (about 6 weeks later), after the field has rotated. For future reference, a field of view in the ecliptic plane hardly rotates at all, but a field of view at high ecliptic latitudes rotates quickly. A rule-of thumb is $1^\circ/{\rm day}$ at the poles, and $0.5^\circ/{\rm day}$ at mid latitudes. It's worth noting again that adding constraints to the observations requires compelling scientific justification, and may make it difficult or even impossible for the observations to be scheduled, so an alternative scenario is to place the map field centers close enough (e.g., such that there is sufficient overlap) that this is not a problem.

**Figure 9.16:** Constraint editor tool.
$\begin{figure} \centering\epsfig{figure=figs4f/consted.ps, width=5.5in} \end{figure}$

There are several different kinds of constraints. ``Timing constraints'' are meant to indicate constraints on specific times, e.g., this comet will be hitting Jupiter on this date and time, and this observation will only be meaningful if obtained at this date and time. These kinds of observations are generally the toughest to schedule, so don't use this unless you really must. This is not what we need here. What we do need are ``Relational Constraints.'' Go to the Spot ``Tools'' menu and select ``Grouping/Follow-On Constraints,'' which pops up a window like Figure 9.16. There are four relational constraint options:

Sequencing, a.k.a. ordered, interruptible group. Sequence must be executed in order specified, but may be interrupted by, e.g., a downlink or calibration sequence.
Chaining, a.k.a. ordered, non-interruptible group. Chain must be executed in order specified and without any kind of interruption. We can't use this here because each of our AORs is already close to the three-hour limit.
Group within. Like a sequence, but order can be shuffled during a time period of specified length.
Follow on. Used to specify scheduling of two AORs separated by a specified time, usually used when the ordering matters. For example, AOR #2 depends on the success or failure of AOR #1, as in we need to make sure we can peak up on a target before exposing for a deep spectrum using IRS. If we had only two AORs, or wanted to break our 4 AORs into two groups, we could use this constraint.

For this project, we want either a sequence or a ``group within.'' We want to be as flexible as possible for ease of scheduling, so we choose ``group within.'' Clicking on that option in the Constraints Editor produces the window in Figure 9.17.

**Figure 9.17:** Group within parameters.
$\begin{figure} \centering\epsfig{figure=figs4f/groupwithin.ps, width=4in} \end{figure}$

By playing with visualization and timing, we learn that a week is about the upper limit for field rotation before left-right (E-W in this orientation) spatial coverage is compromised. We also learn that the overscan in the up-down (N-S here) direction is such that even the most extreme rotation for this field doesn't compromise the coverage, at least not as quickly as in the left-right/E-W direction. So, we enter 7 days, 0 hours into the window in Figure 9.17, and click OK. Then click on the constraint and either drag-and-drop the AORs or highlight each AOR and click on ``get AOR.'' For this example, we put all of the AORs in the same group (see Figure 9.18). We could just put the two North fields in one group and the two South fields in another (could instead use follow-on constraint in that case).

**Figure 9.18:** Filled-out constraints.
$\begin{figure} \centering\epsfig{figure=figs4f/orionv2const.ps, width=5.5in} \end{figure}$

Note: Save your AOR(s) early and often. Now would be good. When you finally go to actually submit your proposal, make sure that you are submitting a clean set of AORs. If you've been playing around with options and alternate targets, make sure to delete those AORs and targets entirely before submitting.

OK, now we're really done with MIPS! The file, ready to be loaded into Spot, containing the raw AORs developed in this chapter (and the IRAC observations in the next chapter) can be downloaded from the SSC webpage:

http://ssc.spitzer.caltech.edu/documents/cookbook/

To develop complementary IRAC observations of the same region, see §10.

Gillian Wilson 2006-11-09