Moonlight observations with H.E.S.S.

July 2022

Previous | Index | Next

A dark night has become a rare sight in most parts of the world. Those who experience it - far away from regions inhabited by humans - quickly realize that the moonless dark night sky in not altogether black. Even if the bright Milky Way is not up above the horizon, the human eye still senses that the night sky is brighter than dark outlines of trees or hills in the foreground. The light results from many sources: large numbers of faint stars, interplanetary dust reflecting solar radiation, and airglow in the high atmosphere all contribute to a sky brightness that is nonetheless a million times fainter than the brightness of the full moon.  The blue skies on a clear day result from sunlight scattered in the atmosphere and the light reflected by the moon during the night is scattered in the atmosphere in the same manner. The night sky during a full moon is hence very much brighter than the night sky when the moon is below the horizon - even if it is 400.000 fainter than the sky during a clear day.

Observations with Cherenkov telescopes such as H.E.S.S. use photomultipliers to register the optical light emitted from showers of elementary particles that are generated in the atmosphere when high-energy gamma rays have crossed the atmosphere to an altitude of about 10 km. Photomultipliers are extremely sensitive in detecting optical photons, even if the photon rate is very low. This allows the detection of the very faint light flashes caused by gamma rays. The faint glow of the dark night sky is much brighter than the signal caused by the gamma rays and the trigger rate of photomultipliers increases significantly when the night sky gets bright due to scattered moonlight. Identifying gamma-ray shower images in such conditions gets more challenging.

Fig. 1: Gamma-ray variability of Quasar 3C279, observed at 100 MeV - 10 GeV gamma rays in 2018 with the Fermi-LAT instrument. Observations with H.E.S.S. (at energies that are 100 times higher) can only be conducted during the pink shaded period when restricted to dark night sky. In this example the benefits of additional observations during moderate moonlight (which was not yet possible with H.E.S.S. in 2018) are apparent from the coverage during the green-shaded periods. Adding a few more nights allows studies of variability during a potentially more interesting period.

In order to record the faint gamma-ray signals in a clear manner and with little confusion, Cherenkov gamma-ray telescopes have traditionally observed only if the sky is truly dark and the moon is either below the horizon or not illuminated (as during the new moon phase).  While this provided data of very high quality, it reduces the amount of observing time compared to other branches of astronomy, which conduct observations even when the moon is up. Irrespective of inevitable periods of cloudy skies, which happen even in the driest and clearest deserts in the world, the limitation to periods of truly dark night sky implies that at most 16% of the 8760 hours of each year can be used for observations.

In addition to the limitation of the total amount of observing time, the constraint to dark night sky also affects studies of transient phenomena. Many gamma-ray sources display variations on time scales of several days (e.g. nova, AGN). Events can be followed without interruptions for 14 subsequent nights at best (for sources that culminate during the night) if observations are restricted to dark night sky. Even a slight extension of this window by adding another three nights of moonlight observation can result in significant improvements of time coverage. This is illustrated in Fig. 1, adapted from an earlier post on the powerful very-high-energy quasar 3C279. When H.E.S.S. operations were extended in 2019, studies of transient phenomena were identified as a subject of high importance. This includes many goals from monitoring and scanning the gamma-ray sky to studying more targets of opportunity when an astrophysical object shows exceptional activity.  The H.E.S.S. collaboration decided to extend operations also into periods of moderate moonlight. Other instruments had already started to perform regular very-high-energy observations during moonlight conditions [1,2] with spectacular success, as shown, e.g., in [3].

Fig. 2a: The sky above the H.E.S.S. telescopes during dark conditions (moon below the horizon) in the first half of a night close to 3rd quarter. This image, obtained by the cloud monitor of ATOM@HESS, is used to derive the atmospheric transparency.

Fig. 2b: The sky above the H.E.S.S. telescopes after moonrise in the second half of the same night. The brightening of the night sky due to moonlight is multiplied by scattered moonlight from light clouds. In this particular example, observations with H.E.S.S. would not be possible anyway due to the high fraction of cloud cover. The large amount of scattered moonlight also increases the night sky brightness to prohibitive levels. The unpredictable distribution of clouds prevents accurate predictions of the night-sky brightness when the moon is above the horizon in all nights that are not perfectly clear.

A dedicated study based on images obtained with the cloud monitor of the ATOM station identified acceptable levels of the night-sky background brightness during those parts of the night when the illuminated moon is above the horizon [4]. The operational procedures for the H.E.S.S. cameras was changed, adjusting the photomultiplier voltages such that observations can be obtained under conditions with varying levels of night-sky brightness. The prediction of night-sky brightness is only possible for truly clear skies. Even very moderate cloud cover has a strong effect, as clouds reflect moonlight very efficiently (see Fig. 2). Fortunately, the skies above H.E.S.S. enjoy a very high fraction of cloudless time. Following these preparations, test observations were conducted in various moonlight conditions and analyses methods were extended to cope with the data obtained during moonlight throughout 2020. In 2021 H.E.S.S. started with moonlight observations during the entire calendar year. Data obtained under moderate amounts of moonlight increased the total amount of observing time by 19% by adding an additional 240h of on-target observations. This new observing mode provided an important contribution to the record amount of data obtained with H.E.S.S. in 2021, when more than 1500h of good quality data were obtained. In particular, the moonlight observations also helped to increase the temporal coverage of variable phenomena, as shown, eg., during the observations of RS Ophiuchus.


[1] M. L. Ahnen et al. "Performance of the MAGIC telescopes under moonlight", APh, 94, 29, 2017.

[2] S. Archambault et al. "Gamma-ray observations under bright moonlight with VERITAS", APh,  91, 34, 2017.

[3] MAGIC Collaboration et al. "Teraelectronvolt emission from the γ-ray burst GRB 190114C", Nature, 575, 455, 2019.

[4] M. Büchele. "Novel Methods for an established System: Moonlight Observations and Deep Learning Data Analysis with H.E.S.S.", PhD thesis, 2021, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany