Meteor observation from space – The Smart Panoramical Optical Sensor (SPOSH)

Detlef Koschny (1), Mario Di Martino (2), Jürgen Oberst (3)

(1)European Space Agency, ESA/ESTEC, Noordwijk, The Netherlands.

(2)Osservatorio Astronomico di Torino, Turin, Italy

(3)Deutsche Forschungsanstalt für Luft- und Raumfahrt, Berlin-Adlershof, Germany

Presented at the International Meteor Conference, 18 – 20 Sep 2003

Abstract

The European Space Agency (ESA) is funding two parallel studies for a ‘Smart Panoramic Optical Head’. The main goal is to develop the technology for a space-qualified, very light-sensitive camera with a wide field of view, both from the hardware and the software side. The scientific application is to allow imaging of phenomena on the dark side of planets or moons, e.g. lightning flashes from thunderstorms or electrical discharges in sand storms, meteors, impact flashes, auroras, etc. This paper will concentrate on the potential of this camera for the study of meteors from an orbit around a planet.

Introduction

As part of a so-called Technology and Research Programme (TRP), the European Space Agency requested industry and institutes to study a camera optimized for the observation of the dark side of a planet or moon. This will require a camera with the following properties:

·space-qualified wide-angle optics;

·Simple, but very light-sensitive detector;

·Software for the detection of the events of scientific interest to reduce the data volume to be downlinked.

The contractors will perform the following activities as part of this study:

·Analyze the science requirements, derive technical requirements

·Produce a design for a space-qualified camera head with software

·Build a ‘breadboard’, that will demonstrate that a flight-qualified version could be built

·Perform a real-sky test

·Deliver the breadboard to the Agency to allow further testing

As part of the “Invitation to Tender” for the proposal, a number of very clear scientific science goals were given. These are listed in the following section.

Detailed science goals

This study is focused on a very clear scientific application: The imaging of the night side of a solar system body. In detail, the following science goals shall be addressed:

(a)Imaging, from orbit, of electric discharges, e.g. lightning flashes, sprites, discharges due to dust storms;

(b)Imaging of noctilucent clouds and aurorae;

(c)Imaging, from orbit, of meteors, i.e. the light emitted when a dust particle enters an atmosphere;

(d)Imaging, from orbit, of space debris entering the Earth’s atmosphere;

(e)Imaging of impacts onto planetary surfaces;

(f)Imaging from a lander that lands on the dark side of a planet or moon.

In the requirements for the proposal, preliminary technical requirements were derived from the scientific goals. As the possible science goals are very diverse, the proposers were requested to concentrate on goals (a) and (c) for the building of the breadboard, even though all science goals have to be assessed in the study.

The preliminary technical requirements – which will be detailed by the proposal teams – are listed in Table 1. They address items such as the sensitivity of the camera, the field of view, how long should it be operating, what is the efficiency of the detection algorithm, how accurate shall the location be identified, how accurate shall the luminosity be identified, and, which spectral sensitivity shall the camera have. Clearly, the main technical drivers are the wide field of view (120 x 120 degrees), which are very large for a space-qualified optics, together with the required light sensitivity. The required light sensitivity means that it may not be enough to design a system with a regular CCD detector, but it may be required to use an image intensifier or a so-called electron-multiplied CCD (addressed in a later section).

Table1: Derived technical requirements for the "Smart Panoramic Sensor Head".

Property

Requirement

Comment

Camera sensitivity

The camera shall Image with a Signal-to-Noise ratio > 5, an object of visual magnitude Mv = 6 mag or brighter, moving with a maximum apparent speed of 5 deg/s

The apparent velocity will affect the dwell time of the object on one pixel.

Field of view

120 deg x 120 deg or larger

Operating cycle

The camera shall record the night side of a planet for more than 80 % of the time.

This affects the relation between exposure time and readout time. Note that if a long exposure is assumed, SCI-020 should still be fulfilled even at the end of the exposure (this might limit the dark current).

Time accuracy

The camera shall allow to determine the time of an event to an accuracy of 10 s.

It is envisaged that this limits the maximum exposure time.

Detection efficiency

The applied software algorithm to detect the events shall not have more than 100 % false detections, i.e. if in one hour there are two meteors above 6 mag, it is acceptable to get two additional false detections. It shall not miss more than 20 % of the events.

This needs to study cosmic ray hits, noise, etc.; which algorithm to use

Spatial resolution

The spatial resolution shall be better than 0.3 deg

The spatial resolution is not so critical.

Dynamical resolution

The dynamical resolution shall be better than 11 bits.

Needed to study the shape of the light curve.

Spectral sensitivity

400 to 850 nm

This range of values shall be assumed for the breadboard. It shall be addressed by analysis what the effort would be to extend the range to 120 nm-- 1500 nm.

Status of the project

At the time of writing this paper (Oct 2003), the scientific studies are almost finished. The contractors have analyzed the scientific requirements as given above and derived more detailed technical requirements to the camera head. It seems feasible to build such a camera, if a space-qualified lens with the wide field of view is developed. Two options are possible for the detector: either the so-called ‘electron-multiplied CCD (EM-CCD)’ which is detailed in the following section, or even a regular, back-illuminated CCD detector. CCD stands for Charged-Coupled Devise and is a standard electronic detector. Back-illuminated detectors are the more expensive version, but the benefit is that they have higher sensitivity than the front-illuminated version.

The dynamical range of>11 bit means that video CCDs are not feasible, as these typically only achieve 8 to 10 bit. Also, their photometric and astrometric accuracy would be smaller.

In space applications, the major resource constraint is the data downlink volume. This is particularly true for interplanetary missions, for example if a camera like this were in Mars orbit. Therefore, a software routine is needed that detects potential events and downlinks only that part of the image that contains the event. This is similar to the well-known software MetRec (Molau 1996 [1]). Some changes need to be done though. Space-based computers typically are slow systems with processor speeds of only a few Megahertz to make them less sensitive to cosmic radiation. Also, flight software needs a number of safety routines that avoid the processor going into an endless loop for example. This software is under development.

The current studies indicate that, from a low Earth orbit, it should be possible to record between 10 and 50 meteors per hour down to a magnitude of 6. For a good analysis of the light curve, the peak magnitude should be around 3 to give enough data points. This would still yield one or more meteors per hour. This should be compared to dust detectors flying in space, typically detecting only one event per day.

For 10 events per hour it is estimated that the camera generates about 250 kBit per hour. Assuming a continuous downlink, this corresponds to 70 bit/s. This has to be compared to the average daily downlink rate for e.g. a Mars mission which is in the order of 2000 bit/s. It can be seen that the meteor camera does not need a significant share of the downlink rate.

The current aim is to obtain up to 10 frames per second imaging rate, however, this may have to be reduced for technical reasons. Definitely it will not be possible to produce video frame rate.

From analyzing available flight hardware, it is estimated that the mass can be as low as 1 kg for the optical head and 2 kg for the complete system.

The main technical challenges will be:

·Develop space-qualified wide-angle optics;

·Ensure the correct selection of the detector – EM-CCD or standard back-illuminated CCD;

·Develop a robust, space-qualified version of the detection software.

The Electron-multiplied camera (EM-CCD)

As mentioned before, one of the possible solutions to the required sensitivity is a so-called electron-multiplied CCD (EM-CCD). This is a fairly recent development (ref xx) and is similar to an image intensifier, but much more robust. On the silicon of a normal CCD detector, a few hundred pixels are added behind the read-out register. These pixels are clocked with a voltage of about 50 V. When the charge corresponding to light signal is clocked at such high voltages through the pixels, it generates an electron-multiplication effect. Thus, at the output amplifier, more electrons per photon are generated than without the electron-multiplication. Thus, the readout noise of the detector becomes less important. The advantage for space applications is that the hardware does not look any different than an ordinary CCD detector and it is a very robust system to fly.

As part of this paper, we have performed tests using an IXON EM-CCD camera. This camera was kindly provided by LOT Oriel from Darmstadt,. Germany. Figure 1 shows a sample image of Lyra, using a commercial 25 mm f/0.85 Fujinon lens, taken under the bright skies of Berlin. Stars of 6.5 mag are still visible at an exposure time of 42 ms.

Comparing this camera to image-intensified cameras, it is clear that this detector would be a promising candidate for the space camera. However, using a large-aperture optics would allow achieving the required sensitivity even with a normal CCD detector operating at 10 frames per second.

Figure 1: Lyra (half hidden behind a tree)recorded with an IXON EM-CCD camera.

Figure 2: The IXON EM-CCD camera.

Potential flight opportunities

This study is proposing to actually fly a camera to observe meteors from space. While the result of the study will ‘only’ be a so-called breadboard, i.e. a demonstration from ground, with non flight-qualified parts, that such a camera can be built, it is clearly of interest to identify potential flight opportunities. These are possibilities:

(a)International Space Station (ISS) – on downward looking platform. Orbit between 300 and 400 km altitude. Operate at night (ca. 40 min per orbit), possibly only above oceans to reduce background illumination

(b)Galileo satellites – the future European navigation system. Several 10 satellites in ca. 23000 km orbit (this would actually need a smaller angle lens which is easier to build)

(c)University satellite in Earth orbit – possibly cooperation with university under study

Ad (a): The ISS orbits the Earth very close to it, only about in 300 to 400 km height. Figure 3 shows the footprint a camera with 120 deg field of view would see. The diameter of the area on ground would be about 1550 km x 1550 km. Of course, operations are hampered by Earth background lights such as cities. These need to be taken into account by the detection software.

Figure 3: Footprint of a camera with 120 deg field of view on the ISS.

Ad (b): The Galileo satellites are the European pendant to the US GPS system, which allows global navigation. The program is just being started, and about 30 satellites are to be launched into space in an orbit just below geostationary orbit. While this is far away, resulting in a loss of signal strength due to the large distance of about 30000 km, it would allow global coverage of the Earth.

Ad (c): To get a camera onto the ISS or a Galileo spacecraft means to participate in a formal bidding process, success is not guaranteed. A number of European universities plan to build small, low-earth orbit spacecraft to demonstrate its feasibility. It may be possible to fly such a camera on a University satellite.

Conclusions

This paper summarizes the current status of an ESA study concerning a “Smart Panoramic Sensor Head – SPOSH”, which may be used to observe the dark side of solar system bodies from space. It shows that it should be possible to build a dedicated meteor camera that could observe the Earth or other planets like Mars and Venus from space and look for meteors.

Acknowledgements

This work is supported by ESA contract no. 17226/03/NL/CP. Thanks to Olaf Koschützke from LOT Oriel for bringing the IXON camera and supporting our tests. Thanks to the Project teams of Officine Galileo, Italy, from Jena Optronik, Germany, and DLR Berlin, Germany.

References

[1] Molau, S., Nitschke., M., Computer-based meteor search, WGN 24, 1996, pp. 119-123.

Property	Requirement	Comment
Camera sensitivity	The camera shall Image with a Signal-to-Noise ratio > 5, an object of visual magnitude Mv = 6 mag or brighter, moving with a maximum apparent speed of 5 deg/s	The apparent velocity will affect the dwell time of the object on one pixel.
Field of view	120 deg x 120 deg or larger
Operating cycle	The camera shall record the night side of a planet for more than 80 % of the time.	This affects the relation between exposure time and readout time. Note that if a long exposure is assumed, SCI-020 should still be fulfilled even at the end of the exposure (this might limit the dark current).
Time accuracy	The camera shall allow to determine the time of an event to an accuracy of 10 s.	It is envisaged that this limits the maximum exposure time.
Detection efficiency	The applied software algorithm to detect the events shall not have more than 100 % false detections, i.e. if in one hour there are two meteors above 6 mag, it is acceptable to get two additional false detections. It shall not miss more than 20 % of the events.	This needs to study cosmic ray hits, noise, etc.; which algorithm to use
Spatial resolution	The spatial resolution shall be better than 0.3 deg	The spatial resolution is not so critical.
Dynamical resolution	The dynamical resolution shall be better than 11 bits.	Needed to study the shape of the light curve.
Spectral sensitivity	400 to 850 nm	This range of values shall be assumed for the breadboard. It shall be addressed by analysis what the effort would be to extend the range to 120 nm-- 1500 nm.