...
- What is histogram of slew times between az,alt pairs?
- Kem asserts it takes about 5 sec to slew between adjacent fields. But that's not right. Assuming we go mostly in elevation, no dome limitation, is angular acceleration limited. We accelerate for half and decelerate for half the time. So the half-slew time obeys (elevation-limited) theta=(1/2)alpha*t^2, and theta = 3.5/2 and alpha for elevation is (conveniently) 3.5 deg/s/s so t=sqrt(2*(theta/alpha)=sqrt(2) = 1.41 sec. So kinematic adjacent-field best slew time is 2.8 sec. Add one sec for settling and we get 3.8 sec, call it 4, not 5.
- ** we should take OpSim output of pointings, with rotator angle, and compute bounding boxes in RA, DEC, then import into GIS database to do rapid queries of coverage in time and filter space.
- Do maximum unobserved gap analysis for each field for each filter.
- Phil Pinto suggests computing next-field time needed as a parallel process.
- Do object-based coverage analysis rather than field-center-based analysis
- How do we represent both transparency and sky brightness across a field? Scalar (what's done now), vector (histogram) or matrix (sky image)?
Some references
LSST science book
...