Important! This is a work in progress. We are gathering constraints here, but these should not yet be considered finalized or comprehensive. There is also an important difference to consider between requirements (especially minimum specs) and expected performance, which could be better than those requirements. In cases where we do not have solid information on expected performance, we are quoting requirements.
For the official requirements on the Telescope & Site subsystems, please see LSE-60 and for the requirements for the Camera subsystem, please see LSE-59. The filter constraints are described in SPT-494.
-
Elevation limits: 20 to 86.5 degrees (with sidereal tracking). These elevation limits are based on constraints on sidereal tracking capabilities. The telescope can point anywhere from 0 to 90 degrees in elevation, but will only track at elevations between 20 to 86.5 degrees.
-
Slew times. There are constraints on the slew capabilities; these are encapsulated in the lsst-ts:ts_observatory_model software package. Roughly speaking, a close-by field requires 5s to slew to, a very distant field could take several minutes. Slews where the elevation changes by more than 9 degrees require a closed optics loop correction which takes an additional 36s (note: current simulations use a 20s delay for the closed optics loop correction and this is still the current requirement; there is a change-control request in progress to update this to 36s).
-
Moon avoidance angle: 30 degrees. The current moon-avoidance angle is 30 degrees, based on scattered light simulations from systems engineering. This is not a physical constraint as much as a 'best practices to avoid scattered light' constraint.
-
Filter change time: 120s. This includes 90s for the filter change mechanism and up to 30s to rotate the camera to align the filter changer with gravity.
-
Number of filter changes. The total number of filter changes must be less than 100,000 over 15 years. This translates to about 18 per night.
-
Available filters: 5. 5 filters can be held in the filter changer at one time. The 6th filter must swap in and out. These filter swaps are expected to be done in daytime, and are limited to 3000 over 15 years, or about once every two days. The filter swap procedure requires at least 1.5 hours.
-
Minimum exposure time: 1s. It is not clear if the PSF is well-formed enough for difference imaging at this exposure time.
-
Maximum exposure time: Undetermined, however at 150s, stars will saturate at about 18.3. At this magnitude, it is unclear how many stars may overlap with existing catalogs for calibration or template purposes, nor is it clear the effect on CR rejection.
-
Sequences of short exposures. Sequences of short exposures (<15s) without breaks (such as a 5s slew) can cause thermal problems for the camera. No more than X exposures of <15s should be taken without a slew or other pause. [??]
- Data Management pipelines may be reconfigured for minisurvey processing, but special processing beyond this is the responsibility of the science community. If your project requires special processing, details of this work should be described in the submitted white paper.
-
The footprint for the WFD must be at least 18,000 sq degrees with a median of 825 visits per field (per SRD requirements). The cadence for these observations is still flexible and rolling cadence options will be considered. (3.4.0.1, tables 22 and 23).
-
Proper motion and parallax requirements (PM accuracy <= 1 mas/yr and parallax accuracy <= 3 mas at r=24) impose some requirements on the overall cadence of the WFD. For example, a sufficient time baseline is required for proper motion measurements, requiring visits to a given field to be spread over many years. (3.4.0.2, table 26).
-
Rapid revisit intervals, uniformly distributed on timescales between 40 seconds and 30 minutes, are required over at least 2000 sq degrees of the survey footprint. Note that these do not have to be consecutive visits, but this area must be sampled over these timescales over the lifetime of the survey. (3.4.0.2, table 25).
-
In the WFD, the single visit r band visit depth distribution requires that no more than 10% of the visits have a limiting magnitude more than 0.3 magnitudes brighter than the dark-sky, zenith r band depth. (3.3.2.1, table 5).
-
In the WFD, no more than 10% of the r and i band visits should have delivered seeing more than 1.1", and the median delivered seeing should be 0.56" (for atmospheric seeing 0.44"), 0.69" (for atmospheric seeing 0.60"), and 0.87" (for atmospheric seeing 0.80").
There are metrics in MAF (called in the run_srd.py script) that will signal if these requirements are met.
-
Deep drilling field positions. 4 of the potential deep drilling fields have been announced; these positions are fixed. The number of remaining deep drilling fields and their cadence of observations is still flexible. In general, these fields are expected to have coadded limiting magnitudes at least one magnitude deeper than the WFD, but even this is loosely specified.
-
Mini-survey observations. The current baseline includes Deep Drilling, North Ecliptic Spur, Galactic Plane, and South Celestial Pole regions as minisurveys. There are many good reasons to include these regions, however their presence is flexible and part of the driver for this call for white papers. Without explicit justification and methods to measure the scientific performance of these mini-surveys, it is likely that their time will be redirected to other needs.
-
It is unlikely that the SRD requirements on the WFD (in particular, covering 18,000 sq deg with a median of 825 visits) can be met with less than 80% of the total survey time. The window of time required for the WFD is thus likely to be between 80-90%, with current simulations using about 86% of the total time for WFD.
-
It is desirable that the WFD reached the SRD-illustrative number of visits and coadded depths per filter:
| u | g | r | i | z | y | |
|---|---|---|---|---|---|---|
| Nvis | 56 | 80 | 184 | 184 | 160 | 160 |
| CoM5 | 26.1 | 27.4 | 27.5 | 26.8 | 26.1 | 24.9 |