Sky quality statistics

There are three main quantities we monitor: Image Quality, Transparency, and Sky Surface Brightness. Much of this metrology comes from our guider cameras which are good, but not great. The photometric calibration accuracy of these cameras is not quite "science grade" so all of these quantities should be taken with a ~10% grain of salt. Still it gives good relative information about sky conditions and the quality you can realistically expect when planning your observations.

Before believing everything you see here, I highly recommend you consider the information contained on the verification page, which describes how these quantities were computed and combined. During normal science operations we use three filters on the guider cameras: SDSS-g` (for VIRUS and LRS2-B observations); SDSS-r` (for LRS2-R observations); and SDSS-i` (for HPF observations). There are systematic wavelength-dependent differences in these sky quality metrics, as expected. The data shown in graphs below have been corrected for these systematic differences and are baselined to represent the i` filter quantities. Further details here.
(produced by iq.py --verify and then iq.py --graphs, after each night has been collected and each year/month has been run)

Image Quality
Sky Transparency
Sky Surface Brightness

Image Quality

There are three measurements of image quality shown on the trimesterly graphs below - the Y-axes are in numbers of minutes.

DIMM - the differential image motion monitor IQ represents the "native site seeing" measured outside of the HET dome (and corrected to be at 55-deg altitude)
WFS - the wavefront sensor IQ represents the seeing obtained from the primary mirror up through the atmosphere, with no impact from segment-to-segment alignment
Guider - the guider camera IQ represents the delivered image quality, measured at the focal plane, similar to what the science instruments receive

Summary points:

DIMM site seeing is worst in T1 (1.2") compared to T2 & T3 (1.0")
median delivered IQ is 1.7-1.8" in T1 and 1.5-1.6" in T2 & T3
requesting seeing less than 1.3-1.5" will significantly limit our chances to observe them, and should be undertaken only with high priority time (P0 or maybe P1)

Image Quality

Transparency

We select guide stars for each observation from the PanSTARRS catalog and therefore know how bright they "should" be. We calculate transparency using the apparent magnitudes and our photometric calibration pipeline for the guider cameras. This calibration is likely reliable around the 10-20% level, so occasional excursions >100% transparency are to be expected.

Summary points:

"Spectroscopic" conditions (50-90% transparency) dominate, accounting for 40-80% of the science time each trimester.
"Photometric" conditions are rare (sometimes only 10% of a trimester) and should be requested only by high priority targets (P0 or P1, maybe P2 if bright time and good availability)
Some "non-spectroscopic" conditions (<50%) transparency are available, and are usually utilized by Priority 4 targets which are particularly bright and can be observed through moderate clouds.

Transparency

Sky Surface Brightness

Again these surface brightness measurements are based on our somewhat rough guider camera calibration pipeline. Whenever a guide star is detected and measured, our pipeline includes a sky subtraction which measures the sky level beyond the edges of the PSF. This is generally reliable but sometimes a double star or poor seeing will artificially brighten the sky level that our pipeline determines. As with the transparency, these sky surface brightness distributions are generally representative but are not high-precision measurements. The guider cameras are small (20"x20") and located at the extreme edge of the pupil.

Another complication is that there are significant differences between g', r', and i' sky brightness levels. These are corrected in a statistical way (described further on the validation page) but this process is non-trivial.

Sky surface brightness