Making an observation with ALIS

Experience has suggested the following steps for configuring ALIS for an observation:

1. Defining a scientific rationale:

This is of course the first and most important task. Naturally it is almost impossible to predict the outcome of any observations. Nevertheless, well planned measurements with a clear view of what needs to be done tend to yield better results than ``ad hoc'' operations.

2. Selecting stations:

Normally it is best to use as many stations as possible, but there might be observation schemes that require only a subset of the stations. It is also important to plan what changes to the rationale are necessary in case some stations cannot take part in the observation (for example due to clouds or technical problems).

3. Selecting imager viewing directions:

This can be either the standard preset viewing-directions (Figure 3.11 and Table 3.6) or specially calculated viewing directions for a particular measurement. It is advisable to prepare a strategy for possible changes if the observed phenomena moves outside of the field-of-view. In most cases it is best to avoid changes of the viewing directions during a measurement. ``Chasing the aurora'' usually results in poor data. Another concern is to make sure that the selected positions are geometrically calibrated against the star background (Section 4.3).

4. Selecting filters:

Filter selection usually follows directly from the scientific rationale. However all filters are not available at all stations (see Tables 3.4 and 3.5). Filter change time increases with the distance between the filters on the filter-wheel (nominally 1-2 s). It might also be necessary to obtain background intensities by exposing images with background filters.

5. Selecting integration times:

The minimum integration time is defined by the shortest exposure resulting in an acceptable $\mathit{SNR}$. As the spectral sensitivity varies for the imager, there is normally a need for longer exposures for shorter wavelengths. Maximum exposure is defined by the longest exposure without saturating the CCD, or, more frequently, by requirements set by the desired temporal resolution. The risk of saturating the CCD must be considered at high binning factors as the sensitivity increases with the binning factors (Figure 3.4). Occasional saturated images pose no threat to the CCD, but yield unusable data. Finding an optimal integration time has proven to require large amounts of experience from earlier measurements. (see also Chapter 3)

6. Selecting temporal resolution:

The temporal resolution is currently limited to one image per 30-60 s at maximum resolution, and is generally never better than 30 images per minute at higher binning factors. These figures vary somewhat between the imagers. Also times for filter changes must be considered.

7. Selecting spatial resolution and sensitivity:

This is generally a compromise between temporal resolution, sensitivity, framerate and disk-space. Note that increasing the temporal resolution by binning might lead to saturation problems due to the increase in sensitivity.

8. Deciding on data storage strategies:

As the disk space at the stations is limited, it is advisable to estimate the amount of data produced and decide what data should be kept, when to retrieve the data and if deleting presumed unusable images should be allowed, etc.

9. Coordination with other instruments:

Such as EISCAT measurements, satellite passes, rocket launches etc. This involves many considerations (for example when does a satellite pass occur? Should ALIS image along the foot-point of the satellite pass?). It is also essential to establish communication with the control centres of participating facilities.

10. Selection of a suitable operational mode for the observation:

Available modes are: automatic, semiautomatic or manual, see Section 5.1.1. In order to increase the number of observations of a particular phenomenon it would be desirable to have more automatic runs, however, experiences indicate that at the present, the best results emerge from manned semiautomatic runs.

11. Defining and testing the required configuration of ALIS:

No usable bug-free program exists. Therefore everything should be tested as thoroughly as possible before running the actual observation. It is also important to start up all systems in ALIS well in advance. The imagers need to be thermally stable and defrosting of the domes as well as booting subsystems at the stations might also take some time. (Estimate at least 1-2 days for testing a new measurement scheme, and 2-4 hours to start up ALIS.) It is also advisable to perform some rudimentary last checkout procedures about one hour prior to starting the measurements.

12. Waiting for the right conditions:

This might be a long wait. In this phase, it is advisable to take images at all stations at regular intervals to check weather conditions and system performance. It is also a good practice to be prepared for many different kinds of observations. For example, if conditions for observing HF-pump enhanced airglow are not right, there might be an excellent auroral event instead.

13. Run the observation:

During measurements many unexpected things can happen, so it is good to be well-prepared with alternatives. Two to three people in the control centre are the ideal for most measurements: one who takes care of technical problems and runs the system, one who concentrates on taking scientific decisions and possibly one who coordinates with other instruments. Taking good notes is essential, as analysis might not occur until months later.

14. Retrieve the raw-data from the stations:

It is time-consuming to retrieve raw-data from the stations. All six stations can be visited in less than a week, but this requires a lot of driving. Usually the disks are brought back only during scheduled maintenance trips to the stations, or as the disks become full. If exceptionally interesting data have been recorded, it might still be desirable to retrieve it as quickly as possible.

15. Analysing and publishing the results:

First one needs to select data, then this data-set must be preprocessed and calibrated before the actual analysis and publication phase starts. These final steps rely on good record-keeping in the previous steps.

Modes of operation

The modes for controlling ALIS are summarised below:

Manual control:

By sending commands directly to ALIS, a user is able to control observations interactively. This mode of operation is most frequently used for checking observing conditions (cloudiness, etc.), for troubleshooting and for maintenance.

Semiautomatic operation:

In this mode a user controls ALIS by running observation control programs (usually these are shell-scripts at the stations) that are uploaded, prepared and tested in advance. This is the normal operating mode.

Automatic operation:

In this mode, an observation control program is loaded and run automatically at a predefined time. This mode can easily be switched to semi-automatic mode if the user wishes to change anything. Exceptions can be reported using a paging system if the control centre is unmanned. Due to the vast amounts of data produced, automatic measurements are uncommon.

Alarms and other exceptions

Many exceptions can be properly acted upon automatically, while others require manual interaction. These exceptions trigger alarms, which are displayed at ALIS-CC and at all user interfaces. If user interface is active, an automatic paging system will also alert the operator on duty. Typically, a pager call occurs for A-Alarms or fatal alarms, (see below) but the paging system can be configured to alert a user when an interesting measurement situation is emerging, etc. ALIS exceptions are classified in three alarm categories:

Fatal alarms:

Fire, trespass or other severe conditions threatening the hardware.

A-Alarms:

Conditions causing the imaging to stop at one or several stations (power failures, filter-wheel problems, unrecoverable run-time errors, etc.).

B-Alarms:

Other conditions affecting the quality of observations (timing problems, communication problems, over- or under-exposed images, etc.).