TECHNIQUES
As an observer, you activity in the OLED project will involve four steps:
Preparation of the occultation. See here.
Capture of the occultation. This is covered below in this section.
Calculation of the light curve. This is covered below in this section.
Submission of the results. Refer to REPORT section.
Some aspects of the UTC timebase of your system will be treated in this section under the heading Time base of your system.
Capture
Depending on your equipment (either video or digital camera), you will be using a specific capture software. For digital cameras popular packages are Sharpcap and Maxim DL, but Genika is also an excellent choice. Images may be recorded in avi (video), ser or fits (digital camera) files. It is important to ensure a correct UTC timestamp in your files, with an accuracy of a few milliseconds. Usually an NPT software like Meinberg's will meet these demands. Other choices may be: GPS antenna connected via USB and with a reliable control software, or camera with built-in GPS. For video cameras IOTA offers a number of possibilities. See Time base of your system below.
Exposure time and gain will have to be adjusted according to the star brightness and background-sky conditions. A rule of thumb is to set the signal to a nominal level of 50% of the sensor dynamic range. Adjustments prior to the event are necessary. For emersions this process is not possible, and we'll have to rely on our experience or use sources of similar brighness for calibration.
Sampling time should be set to the shortest possible time allowed by our system (camera plus PC), without compromising the integrity of the frames (i.e. avoiding missing frames). Sampling times in the range 5 ms to 100 ms are typical.
Light curve
From the images we have to obtain the light curve. The popular Tangra software is more suited for asteroidal occultations, but it can be used for lunar occultations (be warned however that Tangra cannot deal with some difficult lunar occultations). LiMovie is also a nice choice for video. The procedure for obtaining the light curve from a lunar occultation follows techniques that are similar to standard photometric observations, but with important differences, some positive, others negative.
As usual in photometry, the source should not saturate the image.
Control over the spectral band is not essential, as the flux drops to (or jumps from) background level at the occultation in all bands at the same time. In general, photometric filters are not a requirement, and integration over all bands ensures an adequate level of signal from otherwise faint stellar sources in a complicated lunar environment. However, when the source emits strongly in the near infrared, it may be interesting to use a band-pass NIR filter and short sampling times to detect diffraction patterns.
Analyses of emersion and immersion events are similar; we only have to invert the frame sequence when generating the light curve.
The standard technique for obtaining photometry, which is implemented in all the software packages, is the aperture method: three concentric circles are defined centred on the source, the average sky background per pixel is measured in the outer annulus, and the signal within the inner circle is summed, subtracting the background signal. This method will be sufficient for obtaining the light curve of the occultation of bright stars, especially when they are far from the terminator. This method implicitly assumes that the sky background signal is similar to the signal from the dark lunar limb (which is sometimes not true; see here).
Furthermore, in many cases, simply summing the signal within a small circle (a few pixels) around the source (even removing part of the signal from it) may be sufficient to detect the sharp jump in flux and accurately establish the instant of occultation of each star component, even in situations where the terminator (and therefore the illuminated face of the Moon, which can create problems) is closer to the point of contact.
Sometimes photometry will need to be performed in a complicated sky illumination environment, with strong background brightness gradients, especially when the occultation takes place very close to the terminator and the occulted star is faint (which will require adjusting the gain to high values and pushing the exposure upwards). The lunar glow is due to dispersion of light from the Moon in the Earth's atmosphere and the optical elements of our equipment.
This situation may cause the classical aperture-based technique (three concentric rings) to pose problems, as the brightness will not be uniform inside the outer ring and it is not possible to "import" that brightness from another region, which may be subjected to a different light gradient. Therefore, subtracting the average background brightness technique may be counterproductive, and may even introduce spurious distortions, so it may be advisable to simply define a circular aperture and integrate the signal within it: this may be sufficient to detect the "jump" in the light curve, although it will not reflect the actual flux of the source and its detailed evolution over time. This is better than nothing!
Many of the problems mentioned above do not arise for bright stars that clearly stand out from the lunar limb. And, in any case, if they do arise and the objective is not to model the light curve to extract information, their impact is usually not critical (this is the normal case, in which we are only interested in detecting the times at which abrupt changes in received flux occur). For faint stars, however, problems may arise when their brightness is on the order of the brightness of the limb (which can vary greatly depending on sky conditions and, above all, the phase of the Moon and the proximity of the terminator). See here.