4.9. Quality Monitoring System#
The Quality Monitoring System (QMS) provides capabilities to enforce the correctness of the observing process. The purpose of the QMS is thus to:
Ensure that observing programs produced by science investigators during Phase I and II proposal preparation stages do not have errors, meet all requirements for implementation, and do not demand that the system perform in inefficient or unsafe ways.
Watch to ensure that the observing conditions (seeing, sky brightness, etc.) during observation meet the target conditions required by observing programs.
Ensure that the quality of the science data meets target conditions for producing useful science, and that the data are free of abnormal instrumental artifacts.
The above are derived from the Science and System Level Requirements shown here:
Title |
Statement |
|---|---|
Diagnostic Software |
The SWCS shall provide software tools for displaying real- time and long term trends in the performance of individual components/subsystems and to correlate that information with time-stamped data from other subsystems. |
Engineering Data System |
The SWCS shall provide an engineering data system to monitor the health of all subsystems critical to the functioning and performance of the observatory. |
Product Quality Assessment |
The SWCS shall provide software tools to assess the quality of observation data products. |
Monitoring of water vapor, atmospheric seeing, weather conditions, environmental statistics |
The SWCS shall provide the capability to monitor during the night atmospheric turbulence, water vapor, atmospheric seeing, and weather conditions. |
The Quality Monitoring System comprises a server, and visualization and sequencing packages, like other subsystems under the OPS domain. There are three categories that the QMS oversees the management of:
Observations – Observational quality monitoring impacts on-sky data collection. For science the QMS ensures that the observing conditions and observational data meet design criteria given in the science program proposals. It also ensures that alarms triggered during observations do not impact the science data quality. For operations, data quality monitoring feeds information back to the TCS to improve the operations of the AGWS, NGS, and LGS. Observational data monitoring involves such tasks as performing statistical analysis on data to measure: instrument bias and dark current levels, noise and sky background level parameters, the roundness and size of the seeing, saturation of the stars and the science targets, hot and dead pixel statistics and location, instrument background and noise pattern, flux levels of known stars given their known brightness, telescope pointing origin, uniformity of detector response, etc. Part of observational quality monitoring may rely on visual inspection, especially for science data. The Data Processing System provides automated pipelines to extract basic information from images, and the Observing Tools package provides basic tools for quick look analysis, to facilitate data quality monitoring.
Proposals – The proposal system monitors Phase I and Phase II telescope proposals to ensure that they contain proper information and that the program can be implemented. This is especially important for queue or remote observing where entire programs need to be fully complete and ready for execution by the end of the process. The generic monitoring task for both phases is to make sure all the required information fields contain valid content. In Phase I, the system checks to make sure that there is information on targeting, instruments, filters, restrictions in observing window and conditions (sky brightness, seeing, etc.). The system also checks to make sure that the strategies can be implemented, especially with respect to the availability of AGWS and natural guide stars near to their science targets, and that the observing configuration can achieve the Strehl requirement needed for the science goals. In Phase II, observers specify detailed information for observing blocks and on-source integration strategies, including sub-exposures, dithering strategies, mosaic, offsets, etc. The monitoring system ensures that the entries are correct, and that the strategies can be performed without violating guide star, observing window, airmass, and other constraints. In general, proposal quality monitoring is passive, i.e. not involving human operators, but may involve observatory scientists to review and give explicit approval, or to implement more specialized and technically challenging observations.
Subsystems – Subsystems quality monitoring involves monitoring the health of the telescope, instruments, enclosure, facility, and environmental systems. Quality monitoring may either be either passive or active. Active participation by an operator is usually only required for intervention, or analysis after-the-fact during engineering servicing. Under normal operating conditions, system monitoring is mostly transparent to the operator except via visualization displays that summarize the status of various subsystems. The QMS, however, can access data telemetry (data that are time-stamped) from subsystems indicating the overall states of their devices and components. Different types of data processing (Data Processing System) are involved in monitoring telemetry. Most information deemed healthy will simply be recorded into archive without further processing. For time variable and operationally useful or sensitive data such as telemetry for telescope pointing, device temperature, wind speed, seeing profile, etc., the data will be further processed or analyzed to produce visualization or other useful information. The QMS may then compare the results against nominal baselines or trajectories that are established for each device or for the observatory. Information on wind gust and direction may, for example, be integrated graphically into the observational situation awareness display (Figure 4.10). In unusual situations, large drifts in telescope pointing over time may trigger warnings that get logged for further analysis by the daytime engineering crew. The discovery by the monitoring software that guiding corrections need to be made may trigger warnings to update the telescope pointing model, and allow for temporary adjustments.
In normal circumstances, monitoring results are time stamped and stored into the database for future reference. However, in critical circumstances, the results from monitoring analysis may trigger alarms that call for immediate attention. One of the useful concepts of the QMS is the ability to select those subsystems to more closely monitor, and to interactively develop, using graphical tools, on-the-fly analysis procedures. This allows operators the capability to turn any telemetry information of any subsystem or device instantaneously into runtime visualization displays that may be more closely monitored, or that may be correlated with the telemetry of other parameters.