= New project ideas = This is a long laundry list of potential project ideas. Please discuss these with the science operations manager before you invest significant time into any of them -- some may have changed or there may be more important priorities to consider. '''Note: an effort has been made to prioritize, but it's still a matter of judgement. == Relatively low effort, relatively low impact == * TCS/GUI/software projects: * unify all of our sound effects - create TCS event that has a sound file and then make the TCS GUI listen to those events and play those sounds * make list of TO/RA launcher items to remove/updated; share with other TOs/RAs first then send to Chris * {{{vjpg}}}: add color-coded outlines for VHC messages, like on the HTML version of VHC * Queue/science projects: * replicate this, but for HET observations each trimester/year: https://www.youtube.com/watch?v=D5PN8OSyxnE * add TSL keyword for requesting parallel observations (has not been necessary yet) * Image Quality projects: * analyze post-observing measurements to see if they agree w/ WFS-GC gap in IQ - can we account for the IQ difference? * check if guider residuals are correlated with probe positions (e.g., carriage encoder positions); check if they correlate with anything else * look at stored HPFACQ images, estimate focus? accuracy? * Operations projects: * add GPS antenna to DIMM2 celestron while on upper hex anyway? ask celestron? (DIMM2 no longer regularly used, may not be worth doing) * add HPF pedestal level check to daily OPS tasks (SR can add a button) * re-do dark-dome/light-leak test with longer LRS2 exposures, and bay lights; did 5min before, but try maybe 20min? == Relatively high effort, relatively high impact == * TCS/PAS projects: * Improve PAS guider photometry: * allow drawing circular/polygon regions instead of squares * allow sky box drawing * calculate sky surface brightness independent of the stellar profile fit. * add ACQ camera overscan correction to PAS? or flat field? improve real-time processing! * enable subframe readout on HPFACAM camera to speed up HPF setups * Queue/science projects: * after HX is finished, make LST-based over-subscription tool for TACs/users * assume all targets are at center of track, for planning purposes. * estimate pressure and available times of a list of targets / exposures * Integrate with TSL/Hydra * combine TSL parser with visibility calculator tool; add links in hydra, TSL upload to visibility tool * match sky brightness on LRS2 exp calc to real guider measurements * work with Greg Z on (LRS2) telluric standards * prepare "observation report" for PI for each successful target observed at night * single (or multiple) page PDF with sky conditions, guider photometry, WFS details, ACQ setup image, temperature/humidity/etc and other data associated with their observation. Provide on Hydra in real-time. * Image Quality projects: * correlate seeing data w/ gross weather patterns. PCA? * Operations projects: * consider the limitation of 10s "cycle time" requirement when guiding on HPFACQ, waiting to be sure the move goes through? unfortunately slow!! Could we do 6s instead? * analyze DIAFI data from 2.7m on 2022-03-08/09/10 UT (SO working on ACQ data) (can get guider data too) to quantify effect of scattered moonlight in the dome. * what does the pupil see at the edge of a track? anything reflective in the area beyond the mirrors?? PV camera could tell, I think. * check guider statistics - amplitude of corrections, exposure times, S/N, etc; track these on a nightly/per-target basis in the NR, eventually * include HPF software to run on workstations whenever possible == Relatively high effort, relatively low impact == * Consider adding PID loops to WFS corrections - sometimes focus "runs away" and can't keep up * make regular WFS flats with same method as GC, monitor for degradation * get remote control of dome lights, make cal scripts turn on dome lights at end? * show/monitor guider PSFs throughout the night, consider this when focusing and deciding when to stack * make model of physical degradation of M1 based on integrated weather conditions * how much does humidity+dust actually reduce reflectivity? * model or measure it with real data * simulate a whole year with a given closure condition, what is the expected loss of throughput? compare with time spent closed... * subframe readout on ACQ camera (may improve setup times mildly) * get data on dome conditioning. what is effect of poor conditioning on night hours and re-stacks? some weather scatter. drop extreme conditions * add units to TCS/SAMS GUIs for GROC/piston, etc! * ask for units to be added in all GUIs next to any numbers * fix VHC so that it does not rely on outside internet * work-downwind, quantify effects on IQ, verify safety with mech team? then develop formula for wind speed and Az range, auto-calculate (OTT is doing some of this) * include countdown/progress bars on all PAS cameras for exposure times? * include radial profile graphs for all metrology loops (GC/ACQ/HPFACQ?) == Administrative projects to consider == * consider tool for RAs/TOs to provide activity reports at the end of their long OPS shifts or during poor weather, for update on projects * discuss alternative metrics for setup times in evals * right now, are "overhead" and include RA and TO contribs. * if could be just the setup part, would be purely TO (if Niv's stuff works) * think about nossy backups/installations; how much could be run on our workstations? stored in repos? == Other projects to consider == * Ask Steve O if he has seasonal/historic data on sky transparency from 30-inch observations * Work with Greg Z to run LRS2raw reductions in real-time while observing * Refine our pointing model/layers to do a better job for Az=0 and Az=180 - those have the largest scatter in our pointing corrections * Update statistics on sky surface brightness, transparency, and IQ from typical observations at each priority, and provide that documentation to users * use Goldilocks to determine how much sky brightness is "too bright", consider making sky-subtracted S/N measurement in our real-time data reduction * revisit discussion of "throughput drop" from April 2023, investigate causes and possible explanations * monitor total azimuth rotation throughout a whole year - calculate statistics on a nightly or RA-ly basis to see if anybody has a good strategy for managing Az moves * quantify FWHM/IQ between both guiders and WFS PSFs to see what is most reliable * prepare documentation/planning tools to start using Collecting Efficiency * update shuffle to check for proper motion of blind offset starts (has led to bad setups in the past) * write small DIMM data monitoring program, and send audio/visual alert if we go >5min without DIMM data while still getting Guider/etc data * make cronjob status-checker for VIRUS/LRS2 pivot and other data transfer scripts, just to make sure it doesn't die during the night and end up crashing the instrument servers * make monthly probe health check images with cleaned data, to show when flat fields are not working very well * discontinue/remove "rasession" on hpf, if it causes more trouble than it is worth? * add moon/sky brightness to object observability tool, give brightness at start/mid/end of each track. * use ACQ camera photometry to measure photometric zeropoints over many years (requires ACQ data to be processed first - talk to Steve O about his pipeline for doing that) * consider documenting our practice of avoiding observations within 15-deg of the moon, inform PIs about the same, and include this in the planning tools * consider expanding definition of science time to be 12-deg twilights instead of 18-deg as it is now - with HPF (and soon HRS-2) it is possible to observe to -6 or -8 deg twilight. * consider "charging" the -000 engineering program for spectrophotometric standards and telluric standards - right now those are not charged to any partner and are not being consistently handled in our time accounting system * once we are using CE, scan the queue submissions to make sure no targets are requesting impossible/unreasonable CE values given their declinations and exposure times. * someday expand "checkcals" and "plotcals" from only LRS2 to include VIRUS and HPF cals as well * do dark dome testing with GTAG lights on and off some time while closed for weather * add another raspberry PI screen for currently loaded shuffle finding chart * add another raspberry PI screen to show "rasession" on HPF * consider whether this sky model would be useful for us: https://rtn-012.lsst.io/ * consider adding email confirmation when new users create accounts on hydra - would require legit emails and prevent spam user accounts * transition stability_test to Python3 (requires fixing weird issues with errors in the matching code that I didn't write) * consider adding morning Hg+Cd VIRUS cals into an overall morning cal script that will also run HPF morning cals (and any other morning cals needed) in an efficient way * resolve systematic discrepancy between IQ estimates from WFS1 and WFS2 * Someday, include reports from: https://trac.het.astronomy.utexas.edu:8001/trac/WFUCommissioning/wiki/HetDescription/spie and https://www.het.astronomy.utexas.edu/HET/TechReports/techreports.htm intp our document management system: https://edocuments.utexas.edu/D2/#d2 * ask Sergey about adding "-ev" argument for L/V cals, to just run a bunch of things together, V, L, PB, PD. since we do that usually * ask Sergey about adding self-timing feature to "cal" so that it can report how long each calibration set takes; get those numbers and monitor/report/udpate as needed * monitor and update GC1/GC2 positional zeropoints based on Amy W's data * analyze GC2 alignment data from 2022-07-30 UT, see RAlog txt file for details; we moved GC2 across its range, with rich star field throughout. plate-solve and check alignment. it's a very cool dataset!!! * automatically cycle the filter on the inactive guider every ~5min? useful data for sky brightness modeling * consider loading trajectory before cancelling - could speed up the setup commands from target_setup if we run it a little earlier * analyze 2021-11-15 UT data: we intentionally moved a mirror segment out and took a VIRUS SPC observation, trying with multiple segments, using PV camera to verify segment is un-obscured. This would allow a direct measurement of that segment's absolute reflectivity * periodically review HPF evening/morning cals to assess whether anything can be removed or shortened based on what we know now * do another review of HET Queue Operations to publish perhaps in SPIE's JATIS? Similar to this one: https://ui.adsabs.harvard.edu/abs/2007PASP..119..556S/abstract possibly starting from July 2016 when full science operations began. Maybe target July 2026 for 10-year review? * add longcals to highpri, to help remind us when longcals are needed? * update hetsky page, migrate off that computer, improve maps/links/etc * periodically verify probe positions using VIRUS + ACQ + Guider1 + Guider2 data in rich field - plate solve everything and map it all together * data mine historical weather data, how does the temp change before sunset, seasonally/weather; use that to inform operational decisions *