Skip to main content

ASPA: the making of (personal account by Emmanuel Dekemper)


The ASPA instrument had quite an unconventional birth compared to the usual remote sensing instruments that we develop at BIRA-IASB. It was designed, built, tested, and used without any dedicated budget line. Still, to my experience, its realization is a huge success given the limited resources (both in time and money) that were available. And this could only happen thanks to the motivation and commitment of the engineering department, which backed the ideas of some scientists to turn a nice concept on paper into a working and reliable instrument, operated in a harsh environment.

Since several years now, my physicist colleagues Gael Cessateur and Hervé Lamy were trying to measure a small feature of the Northern lights: a subtle polarization which had recently been discovered by a team of scientists from the University of Grenoble. The fact that some auroral spectral lines are polarized and some others are not is still puzzling the scientific community, which is calling for more experimental evidence.

Gael and Hervé had developed an instrument, Premier-Cru, whose concept was inspired by the French instrument that first detected this polarization. As this instrument never really delivered satisfactory results, they were open to new instrumental concepts capable of accurately measuring the polarization of auroras.

Interdisciplinary science with AOTF filters

I was totally ignorant about the physics of auroras, and my area of expertise was more into the remote sensing of atmospheric trace gases using optical instruments. The project I’m most involved in, namely ALTIUS, is a satellite instrument whose original concept is to use special spectral filters, called AOTFs, in order to analyze the absorption of natural light (Sun, stars) by the atmospheric molecules. Those filters work under the principle of light and sound interaction inside a transparent crystal of tellurium dioxide (TeO2). The fundamental physics of this interaction was developed by the French physicist Leon Brillouin in the 1920’s.

Actually, ALTIUS is not the first space instrument using AOTFs, neither was it the first time for BIRA-IASB to dive into this filter technology. Indeed, AOTFs have been used in planetary missions such as Venus Express, or ExoMars. No need to say that there was quite a lot of expertise in the Institute both in terms of controlling these devices, and understanding their behaviour.

The ALTIUS concept has also triggered a nice ground-based application: the NO2 camera. Originally a prototype of the visible channel of ALTIUS, I made it evolve into a field instrument capable of mapping the abundance of NO2, a harmful polluting species generated in combustion processes, above power plants, or cities. The NO2 camera is also based on an AOTF to take spectral images of the scene under investigation.

Exchanging with Gael and Hervé from time to time, we realized, bit by bit, that AOTFs could actually be the core of a new instrument dedicated to the measurement of the polarization of auroras. Indeed, one feature of the AOTFs is that they physically separate the incoming light into two components of perpendicular polarization. This feature is usually used to “clean” the light after crossing the AOTFs, keeping only one polarization for further analysis. But here, we could use it to its full potential.

ASPA had to grow up quickly

The real kick-off for ASPA took place in November 2019. This is when I started designing the instrumental concept, collecting opto-mechanical parts datasheets, and evaluating the performance of the future instrument.

Figure 1: First opto-mechanical design of ASPA.

Soon enough, I realized that auroras are not always the bright curtains dancing in the sky that we all know from documentaries. I had to be sure that the concept would be sensitive enough, at least down to some realistic brightness levels. Affordable detectors allowing to detect very faint signals are not really common, such that I had to work out the optical design to ensure that photons would not be lost on their way through the AOTF and to the detector. The crucial point was the focusing of the light beams onto the smallest possible amount of pixels, based on affordable optical elements, of course.

Aspa diagram
Figure 2: Back-end optical layout of ASPA: a pupil is defining the field of view of the instrument, the green pieces are the lenses, and the yellow object represents the AOTF. The blue and pink beams represent the two polarization states.

The manufacturing of the instrument turned out to be a race against the clock! Indeed, Hervé was planning to join two French teams by the end of February in Skibotn, in the North of Norway. The goal was to have ASPA there and ready to see its first auroras.

Figure 4: The CAD (computer-aided design) model of ASPA.

In December, engineers Sophie Berkenbosch and Jurgen Vanhamel had already started conceiving the electronic chain driving the AOTFs of ASPA. In January, most opto-mechanical parts had been purchased and delivered to BIRA-IASB, and the assembling of the elements could start. In the meantime, Claudio Queirolo (B.USOC operator and engineer) had produced a CAD (computer-aided design) model of the instrument, allowing to have a definitive view on the dimensions of the beast. This was needed to move on with the organization of all the different parts into a spare box we had from a different instrument. Engineer Pepijn Cardoen then took charge of the accommodation of all the parts into the housing: RF (radio-frequency) drivers, amplifiers, thermal sensors, heaters, inclinometers and fans found a cosy place next to the optical modules.

after integration
Figure 5: ASPA after integration. Credit: Gaël Cessateur.

In parallel to the hardware work, engineer Roland Clairquin developed a control and acquisition software from scratch, making sure that all operating needs would be fulfilled in his code. During the first week of February, a first version of the software was available, and the instrument could go through a minimum set of laboratory tests. To my great relief, everything worked surprisingly well: the code was almost bug-free, the alignment of the optical elements was matching the drawings, and the behaviour of the AOTFs was just as in my computer model! The only thing that we couldn’t test was the operations in the campaign conditions.

And then, a week before the campaign, when the instrument and all the equipment was packed and ready to be shipped to Norway, trouble started…

In order to get our equipment in Norway, which lies outside Europe (in the world of border control and tariffs at least), we needed to get the “ATA carnet” from the Chamber of Commerce. Not the fastest procedure. It took a week and 350 EUR to receive it. We got it just a few days before we left, and the only shipping option left by then was DHL express: “24h delivery wherever in the world”. Well, forget about that!


The DHL adventure

First, it started with a DHL employee coming to pick up the parcels, and realizing he couldn’t take them because one of them was heavier than 30kg, something which was clearly indicated on the shipping order. Second, after a truck came to pick up our equipment, we received a first phone call asking for a pro forma invoice. Something you shouldn’t need when you have an ATA carnet. The day after, on Friday, it turned out they had lost the ATA carnet… but recovered it later that afternoon, only after the customs offices had closed for the weekend.

On Monday, the three parcels were still at DHL Brussels, and they forced us to produce a fake pro forma invoice in order to pass the customs. Due to customs closing at noon, they finally scheduled the shipment for the day after. On Tuesday, the three parcels finally passed customs, ready to be shipped during the night…but on Wednesday, it turned out that one of the three parcels (the one containing the RF equipment) was on its way to… wait for it… Madrid! Despite this, we decided to drive back from Skibotn to Tromsø to pick up the two parcels that had arrived, and pay the University of Tromsø a visit to find some replacement RF equipment. Thanks to the kind collaboration from the University of Tromsø, we were already able to re-build the instrument on Wednesday, and make it work, though not in ideal conditions.

Figure 6: Busy re-assembling the instrument in the bungalow on Wednesday evening.

On Thursday afternoon, the last parcel finally arrived in Tromsø. We went back to Tromsø to pick it up, each travel back and forth between Skibotn and Tromsø costing us a lot of time, fuel, and stress (driving hundreds of kilometers around the fjords, in winter conditions beyond the polar circle, is something Belgians are not used to). Ten days after the campaign, we received the bill from DHL (without any discount for the substandard services).

Icy Roads
Figure 7: The driving conditions between Skibotn and Tromsø...

A happy end for ASPA?

Luckily, we didn’t miss any good opportunities to observe auroras while being stuck in limbo with the shipping of ASPA. On the flip side, we didn’t get any decent auroras during the last three nights of observation either. Only very few occurred in good visibility conditions (no clouds), and they were all extremely faint and definitely below the detection limit of ASPA: while the typical green auroras can be as bright as hundreds of kiloRayleigh (a unit to measure the brightness of airglow and auroras), the ones we saw remained below 1 kiloRayleigh, which was our ultimate detection limit.

What I take home from this endeavour is the wonderful commitment of many talented people from BIRA-IASB. Together, we managed to design, manufacture, and test a completely new instrument in barely two months! I am of course a little disappointed, because of the few and faint auroras we were given to observe during the three nights of observation with ASPA. I would have loved to be able to return with nice results, to show everyone it was worth working hard to be ready for the campaign, but nature  (as well as DHL, apparently) is unpredictable. On the bright side:  ASPA is on location and ready now. We will make a second attempt this fall until we get good observation conditions!

Figure 8: Setting up ASPA for a night of observation in the dome of the observatory of Skibotn.
Figure 9: One of the few auroras we saw during the campaign. Don't be misled by this picture: with the naked eye, the aurora were only slightly noticeable.


News image 1
News image legend 1
Aurora in the Norwegian Lofoten. Credit: Jeroen van Gent.
News image 2
News image legend 2
Figure 1: First opto-mechanical design of ASPA.
News image 3
News image legend 3
Figure 3: The red cross indicates the location of the observatory where the observation campaign took place, nearby Skibotn.