Temperature measurements of the OH layer in the upper middle

atmosphere during FESTA-campaign on Svalbard


Florian Geyer, Ulrich Hamann, Erik Rodesjö, Paula Sankelo , Åsmund

Skjæveland, Joakim Skytte, Teresa Tenhunen and Hanna Tietäväinen



Middle Polar Atmosphere, AGF-210, FESTA-campaign report



Table of content

Table of content 1

Abstract 2

1. Introduction. 3

2. Physical conditions in the upper middle atmosphere (70-100 km) 3

3. Instrumentations. 4

3.1 All-Sky Video Camera (ALSC) 4

3.2. Meridian scanning photometer (MSP) 6

3.3. Ebert-Fastie Spectrometers. 6

3.4. LIDAR.. 9

3.5 EISCAT Svalbard Radar 11

3.6. Meteor radar 11

3.7. Rocket instrumentation. 12

4. Results and discussion. 17

4.1. All Sky Camera. 17

4.2. MSP. 18

4.3. Spectrometer 21

4.4. Lidar 30

4.5 Eiscat 32

4.6. Meteor Radar 34

5. Conclusions. 44

6. References. 44

Appendices. 45

Appendix A.. 45

“One Meter Green” Spectrometer – Cookbook / Instruction manual 45

Appendix B.. 46

“Silver Bullet” Spectrometer – Cookbook / Instruction manual 46

    Appendix C.. 47

MSP (Meridian Scanning Photometers) – Cookbook / Instruction manual 47

Appendix D.. 49

All Sky Video Camera (ALSC) – Cookbook / Instruction manual 49

Appendix E - Timetable. 49

Appendix F. 50



Temperature measurements of the OH layer in the upper middle atmosphere were conducted from 4th of November to 15th of November 2002 on Svalbard. These measurements were part of university level campaign called FESTA. OH temperatures were obtained by one meter and half-meter focal length Ebert-Fastie spectrometers and LIDAR measured the height of OH layer. Comparative work of OH temperatures and heights could be done with meteor radar and EISCAT measurements. For discovering the interesting and best observing periods the MSP (Meridian Scanning Photometer) and ALSC (All-Sky Camera) were used. Because this kind of large measuring campaign with so many instruments running simultaneously have never been done before, there were some anticipations for new results and these hopes were partly fulfilled.

1. Introduction

FESTA (Focused Experimental and theoretical STudent Activity) is a campaign included in UNIS (University courses on Svalbard) course The Middle Polar Atmosphere, AGF-210. Campaign is planned being run every year but presumably with different content. This year FESTA was about observing and measuring the conditions, especially the temperature of so called OH-layer, in the upper middle atmosphere. The campaign was carried out on Svalbard at 78°N. Temperature measurements of the upper middle atmosphere are quite rare at such high latitudes, and so Svalbard with its comprehensive collection of different instruments plays quite an important role when it comes to these kinds of campaigns. Dark polar night in winter makes it possible to have long measuring periods with light-sensitive instruments. Auroral activity on Svalbard, as in polar regions, is generally high and can complicate especially the OH temperature measurements.


Many various instruments in different locations on Svalbard were run together during this campaign, some of them were running on a routine basis and some especially for this campaign’s purpose. The one meter focal length spectrometer was placed at Auroral Station, near Longyearbyen, together with MSP and ALSC, while the other spectrometer as well as LIDAR were situated in Ny-Ålesund, approx. 120 km from Longyearbyen. EISCAT and meteor radar are placed only 5 kilometers away from Auroral Station.


2. Physical conditions in the upper middle atmosphere (70-100 km)

Temperature reaches its coldest values in the upper part of the mesosphere, being approximately 200 K in the mesopause at the altitude of 85 km. After this, i.e. in thermosphere, temperature begins to rise again with increasing altitude. Major constituents in upper mesosphere and lower thermosphere are N2, O2 and H2O. Especially water vapor plays important role giving with sufficiently low temperatures good basis for noctilucent cloud formation in summer and also for the airglow phenomena in OH-layer at altitude about 83 km.


Major sources of heat are provided by ionization of atoms and molecules by absorption of extreme UV-radiation while radiative cooling through emissions is associated with the vibrational relaxation of CO2, H2O and O3. Many interactions in the mesosphere and thermosphere are caused by collisions of energetic electrons with various atmospheric constituents. These collisions are an important part of the chemistry in the upper atmosphere. Also various atmospheric waves carrying significant energy, for example diurnal and semi-diurnal tides forced by solar heating, are observed in the upper middle atmosphere.


This is also the layer where meteorites usually burn up as they enter the atmosphere. That is why it is possible to use meteor radar data in comparison when trying to find out the temperature of this layer. With various remote sensing instruments, spectrometers, photometers, resonance fluorescence detectors and rockets can the characteristics of the upper mesosphere and lower thermosphere be studied.

3. Instrumentations

3.1 All-Sky Video Camera (ALSC)

All-sky video camera consists basically of a fish eye lens, a light intensifier, relay optics and a video camera. It monitors the total sky in visible wavelength region. The time resolution of ALSC is 25 frames per sec. and spatial resolution 0,5 degree near the zenith.



Figure 3.1.1: Working scheme of the ALSC


All-sky video camera is used to monitor prevailing conditions in sky. The real time all-sky images can be watched from the monitors in the computer room and the kitchen of Auroral Station and they are also updated to Internet on a web-page that can be found through the home-page of the station (http://haldde.unis.no/).


The all-sky images obtained in real time is used for example to get an overview of the auroral activity. Below is a image of sufficiently strong aurora covering most of the sky.



Figure 3.1.2: All-sky video frame from the Auroral Station in Adventdalen, Norway, 6th of November 2002.


Because the purpose of this campaign was to investigate the conditions in the upper middle atmosphere, and since auroral activity can cause problems in that, the interesting periods were the ones with clear sky but no aurora. Light pollution, especially on cloudy days, is an outstanding problem with the all-sky camera.



3.2. Meridian scanning photometer (MSP)

The light from the sky is reflected into the MSP by a rotating mirror mounted in a 45-degree angle to the ground. It’s rotation frequency is one rotation per 4s. The scanning is along the magnetic meridian from north to south. Then the light is distributed to 5 detectors, which measure intensity on the wavelength: 5577Å(green aurora), 6300Å(oxygen, red aurora), 4278Å(molecular nitrogen, blue aurora), 4161Å(hydrogen, protons), and 8446Å(atomic oxygen), which are called channel 1 to 5 in the following. Narrow band interference filters with a bandwidth of 5Å filter out the measured wavelength. The light beam passes a photomultiplier and is detected by a counter.


The filtered wavelength is dependent on the tilt of the interference filter. By tilting, the optical way between the to mirrors is increased and the transmitted wavelength increases also.



The aurora station always receives radiation, which has peaks in already known wavelength, also the wavelength the MSP measures. The filters were slowly tilted and the transmitted intensity is measured. An angle is recorded at the maximum transmission (peak). Another angle is recorded near the peak where the transmission is not due to the peak but due to the background radiation. The difference between them is the signal.


For channel 1, 3, and 5 the calibration with the light from the sky is always possible. For channel 2 and 4 a strong aurora is needed.


Figure 3.2.1: MSP central unit


3.3. Ebert-Fastie Spectrometers
Spectrometers measure the intensity of incoming light as a function of wavelength. Unwanted wavelengths are first filtered out by applying a bandpass filter. The instrument will scan through the relevant wavelength region and create a spectrum by measuring ten spectra and taking a running mean.

Light enters the instrument through a slit placed in the focal point of the concave mirror. All incoming light will thus be reflected so that the reflected rays travel towards the grating parallel with each other. Reflection angle from the grating depends on the wavelength in a manner given by the grating equation. Light reflected from the grating will again be reflected from the concave mirror, and a specific wavelength will arrive in the other focal point, where the exit slit is placed. After the exit slit light goes to a photomultiplier and the intensity is measured as counts. As the grating is being turned, different wavelengths will hit the focal point and a spectrum is obtained. Figure 3.3.1 shows the optical elements of spectrometer and figure 3.3.2 a diagram of the path of light. Following numbering applies to both figures:

  1. Entrance slit
  2. Concave mirror
  3. Grating
  4. Exit slit
  5. Lens
  6. Detector
  7. Band pass filter

         Figure 3.3.1: Optical elements of the spectrometer


Figure 3.3.2: Diagram of the path of light in spectrometer


“One Meter Green”

One Meter Green is used for measuring blue light from the proton aurora, and for this reason it should be pointed along the magnetic field line, here 8° south. Field-of-view of the instrument is 8° by 8° and time resolution from eight seconds to five minutes. Spectral resolution is from 1 Å upwards, depending on the width of the slit, which can be adjusted. ((In the setup being used here the width of the slit is 1 mm, which gives a spectral resolution of 4 Å.))

Grating is being turned by a stepper motor, so in the scanning procedure backlash has to be taken into account. After each scan the grating is being turned back until it has passed the home position, and then taken into home position again. Thus the backlash is removed before scanning again.  

“Silver Bullet”
Silver Bullet measures near-infrared light emitted mostly by hydroxyl and oxygen in the airglow layer. It is pointed towards the zenith and has a spatial resolution of 5°. Time resolution is from five minutes upwards and spectral resolution is 4 Å. Scanning is done in a quite different manner from One Meter Green, as the arm that moves the grating is resting on a wheel that slowly turns. The wheel is shaped so that the grating falls into home position once in every turn, and then scans over the relevant wavelength region before falling back into home position again. This way there is no backlash to be removed.

“Half Meter Black”
The Half Meter Black Spectrometer in Ny-Ålesund is working in the same way as the Silver Bullet Spectrometer at Auroral Station. It is pointed towards zenith and measures near-infrared light between 7250Å and 7450Å.


3.4. LIDAR

The LIDAR (LIght Detection And Ranging) in Ny-Ålesund is a useful tool for remote sensing. For this campaign LIDAR measured the height of the OH layer. It can also determine vertical profiles of ozone and aerosols in the atmosphere.


The LIDAR can give a good image of the height of OH concentrations in the upper part of the middle atmosphere. OH is being produced by dissociation and recombination of different atoms and molecules. As the laser beam from the LIDAR hits the OH molecule, the energy in the light excites electrons. As they fall back into equilibrium position, light is being reemitted. Due to the molecule’s characteristic vibration and rotational properties the light is being emitted like a lighthouse. Since every molecule emits in a unique way, the OH (and thus it’s height) is easily detected. 

(See fig. 3.4.1).





Text Box: Figure 3.4.1: LIDAR data from 6th of November 2002

We usually refer to this as the light is being backscattered by the molecules. Height is being determined from the transit time of the light pulses and variation in strength gives the profile. In this case however, the interesting part of this whole profile is the specific height of the OH layer.


The LIDAR’s main components are a Neodyn-Jag laser

and a receiving mirror. In addition a trigger sets of the laser. A counter registers the backscattered light received by the mirror and finally, a computer processes the data (Fig. 3) There are also a number of other components that is needed in order to make the LIDAR functional (Fig. 2).


The image to the left shows the laser and receiving mirror of the LIDAR in Ny-Ålesund. The fact that they are so closely placed makes it impossible to use it at very low altitudes. However the LIDAR ranges from approximately 8-120 km[1], from the lower to upper middle atmosphere.









          Backscattered light  






                                                     Trigger                               Counter





3.5 EISCAT Svalbard Radar

The EISCAT radars on Gruve 7 mountain are incoherent scattering radars designed for studying the ionosphere. One 42-m diameter dish is fixed pointing along the magnetic field lines. The other, a 32-m dish, is fully steerable. Both operate at 500 MHz, with a peak power of 1 MW. The radars are primarily intended for studying the upper atmosphere from 90 kilometers and up, but they can also be used to study lower altitudes, down to approximately 80 kilometers. However, then the signal suffers from strong ground clutter, due to the clear view of the sea and of distant mountains.


The 32-m diameter radar antenna was operational in 1996 at 0.5 MW peak power. The 42-m antenna was added later, and the facility has been upgraded to 1 MW peak power. The antennas are normally operated simultaneously, with the beam switching from one antenna to the other several times each minute.


Figure 3.5.1: The 32-meter antenna


In this campaign, the radar was used to get temperatures at around 85 kilometers altitude. The dish was pointed in a low elevation (36°), so that the beam crossed the mesopause roughly over Ny-Ålesund. This had the advantages that ground clutter was reduced significantly, and the temperature data for the atmosphere over Ny-Ålesund was obtained, which is also where we have LIDAR OH profiles. On the other hand, the data quality was reduced because the beam passed through much more atmosphere, and contained a lot of information on horizontal gradients as well as on the vertical electron density, thus made it harder to fit the measurements to a model to get the temperatures.


The radar cannot get the neutral temperature directly, it can only get ion and electron temperatures. But the collision frequency around 85 kilometers is high enough to assume that ion, neutral and electron temperatures are the same, as long as there is no aurora or other disturbances that bring the atmosphere out of thermal equilibrium. Unfortunately, the D-layer is often very weakly ionized when there is no such disturbance, so it was not certain that useful data from the quiet periods, which were most interesting, had been obtained.



3.6. Meteor radar

The Svalbard meteor radar is owned by National Institute of Polar Research (NIPR), Japan, and operated by University of Tromsø. It is situated at the foot of Gruve 7-mountain, where it was installed in March 2001. The radar field consists of six broadband receivers and a transmitter that is operating at frequency of 31 MHz and 7.5 kW peak power. Meteor trail measurements are being made throughout the year and around the clock, independent of weather or light conditions.  


When meteors hit the Earth’s atmosphere and burn up, the heat ionizes neutral constituents of the atmosphere. These trails of ionized particles produce radar echoes, which can be used to derive the wind speed and temperature at the location of the trail. Trails disappear quickly due to diffusion, and diffusion speed is directly proportional to temperature. Wind speeds are inferred by the Doppler shifts of the radar echo. Trail production is at its largest between 70 and 100 kilometers, which is just the region of interest of the FESTA campaign. Unluckily the variation in temperature measurements is very high in altitudes from 75 up to 95 kilometers, and thus the derived temperatures are not reliable for low altitudes. One of the suggested goals of the campaign was to study how the mobility parameter and therefore also the diffusion coefficient should be varied in order to produce more reasonable temperatures.




Figure 3.6.1: Nippon / Norway Meteor  Radar in Svalbard


3.7. Rocket instrumentation

Mini Dusty is a project of University of Tromsø (UiTø) and Norwegian Defence Reseach Establishment (FFI). They have developed miniaturized payloads for measuring mesospheric parameters. The payloads have been launched up on a small 150 G Viper IIIA rocket motor in Andøya Rocket Range until year 2000. It was the intention to use rocket measurements as a part of this FESTA campaign, but unfortunately it was not possible. The new version of the motor, Excalibur 2, is being tested at the time and will presumably be in use in 2003. However, the rocket instrumentation was studied during the campaign.


A payload weights 7.5 kg and the outer diameter of the payload is 5.4 cm. It contains dust, electron and ion probes, an ozone photometer, a temperature probe and a Faraday probe for measuring rotation. The probes are connected to a motherboard and a slaveboard. The rest of the electronic part contains a battery, a power control and supply, pyro electronics, a transmitter and an antenna in the tail section. The payload is equipped with an ejectable nosecone and a heat insulation tube with aluminum foil.


There are four electrometers on the motherboard and four on the slaveboard. They measure the electric current produced by the different probes. The one that is furthest from the digital circuits is used as the most sensitive one since they can affect the electrometers. All the electrometers have a sensitivity of  1 × 10-13 A and each of them can be set with a reference voltage from –7 V to +7 V depending on the connected probe.


All the eight electrometers, as well as the other instruments on the payload, are calibrated as close to the launch as possible. The payload should let run for at least 15 minutes and an external current source for an hour for stabilization before calibrating the electrometers. When the current source is set to zero the offset is adjusted until it reads zero. If there is too much noise the payload and electrometers must be shielded better. The applied current can be checked on the Eidel display. The gain needs to be adjusted until the applied current value is equal to the measured one. This is done with both positive and negative currents. The electrometer offset is rechecked and adjusted if necessary.


The current source can be connected to a computer. A program is used to run a current sweep of the whole current range for the electrometer. When calibrating is done correctly the current line reaches the applied current both in negative and positive sides and the zero current is set to the zero level.


In Figures 3.7.1 and 3.7.2 the electrometer is calibrated correctly. There was a 2.0 × 10-7 A applied which was reduced by a factor of ten for each sweep. Figure 3.7.1 shows all the sweeps while Figure 3.7.2 shows the fourth sweep with maximum current of  2.0 × 10-10 A. Figures 3.7.3 and 3.7.4 are examples of uncalibrated electrometers. There was the same current used.






















Figure 3.7.1. Correctly calibrated electrometer



















Figure 3.7.2. Correctly calibrated electrometer. The fourth sweep.


















Figure 3.7.3. Uncalibrated electrometer





















Figure 3.7.4. Uncalibrated electrometer. The fourth sweep.


The rockets are the only instruments giving in situ measurements from the middle atmosphere. The height resolution can be down to tens of centimeters. Together with ground-based instruments, such as radars, lidars and spectrometers, rockets can obtain comprehensive information from the middle atmosphere phenomena.



4. Results and discussion

4.1. All Sky Camera

Light pollution in Adventdalen

Light pollution is an increasing problem at the Auroral station. Longyearbyen is growing, and is expanding toward Adventfjorden and Adventdalen. A picture from the all-sky camera easily illustrates the situation: Light pollution from the settlement is the dominating feature in the image. When the sky is clear, the problem is not as obvious, but light from the town will be scattered around and into the camera also on a clear night. Cars and trucks passing on the nearby road are also a problem, and show up in the MSP and ASC data. This light pollution problem is the main reason for the current plans to move the station to the EISCAT site, where it will be above much of the light pollution.


Figure 4.1.1: The light pollution from LYR an overcast night

Figure 4.1.2: Clear early morning sky with some aurora

Figure 4.1.3: A clear night with a passing car

The figures show the light pollution in different situations. In the first, the light is scattered by low clouds and dominates the image. In the two other pictures, the sky is fairly clear and bright celestial objects can be seen, but the town can still be seen in the top of the image. The bright spot near the bottom is town light reflecting off another dome.


The light pollution is not only interfering with the all-sky camera. When it was overcast, a strong nitrogen line showed up in the spectra from both the spectroscopes, and the source is Longyearbyen’s streetlights. The half-meter spectrometer in Ny-Ålesund has a non-linear background that is probably related to light pollution.


 Light pollution was less obvious in the MSP data, which appears to be generally slightly brighter to the south. However, this may be an effect of scattered sunlight, since the sun is near magnetic south in early morning hours.


At around 06 UT, the twilight from the not quite rising sun is much brighter than the city and the all-sky camera shows very bright image.


4.2. MSP

In order to validate the data collected, we must be able to proven the conditions in the atmosphere are what we claim them to be. For example, there can be not auroras in the region where the temperature measurements are being conducted.


The MSP combined with the ASC gives a good image of the aurora activity.

While the ASC delivers visual confirmation of aurora activity, the MSP provides us with information such as type and strength of the atmospheric emissions. The aurora signal itself may be used to determine the kind of aurora, but in this context it is only used to make sure that the temperature measurement is not disturbed.


Unfortunately weather in the middle atmosphere was not applicative for the measurement of this campaign during observation time. In fact only during three nights atmosphere was sufficiently quiet to get good readings from the spectrometers. The disturbances were mostly overcast or heavy particle precipitation from space. However, other instruments didn’t experience the same to be conditions to be equally bad, for example EISCAT.


The 4th channel (Hydrogen) has to be calibrated by use of an artificial lamp or strong aurora. This was not done and therefore the data of this channel has a larger uncertainty than the other channels (one can still se the pattern following that of the other channels though).


The intensity of the incoming radiation at the 5 observed wavelength is plotted color coded in dependence of the time and projected latitude at the magnetic meridian.

As the graphs below (Fig. 4.2.1, 4.2.2, 4.2.3 and 4.2.4) show, the atmosphere was quiet during the 5th, 6th and 11th of November.

Figure 4.2.2: Keogram of November 5th 2002, Longyearbyen


Figure 4.2.3: Keogram of November 6th 2002, Longyearbyen

Figure 4.2.4: Keogram of November 12th 2002, Longyearbyen


Figure 4.2.5: Keogram of November 11th 2002, Longyearbyen


The black fields represent periods when the MSP wasn’t running, mainly due to daylight.

What we are interested in is periods with no aurora activity in the zenith angle, 90º, since that is where the LIDAR and the spectrometers are looking. During both nights however, some auroras where sweeping over the sky (shown as green and red bands from top to bottom).


Figures 4.2.5. and 4.2.6 present the aurora data of morning and evening of November 5th 2002. As the graph show, almost no auroras occurred during Tuesday from 00.00 to 04.00 UTC, from 16.00 to 24.00 UTC. We had some instrumental noise showing as lines from 0 to 180º.


The gap from 19.36 to 20.36 on the evening of the 5th was a result of some operational procedures that required shutdown of the MSP.  While the gap from 18.00 to 18.30 UTC on the 6th, was due corrections of some instrumental error.


A whole timetable of running time of the MSP and periods without aurora is shown in the appendix.



Figure 4.2.6: Keogram, morning of November 5th 2002, Longyearbyen


Figure 4.2.7: Keogram, evening of November 5th 2002, Longyearbyen

4.3. Spectrometer

Three different spectrometers were used to obtain data from the mesopause region between 80 and 90km height. The aim was to calculate temperatures in this region using emission lines from OH radicals (the so-called airglow phenomenon).


The Silver Bullet Spectrometer in the Auroral Station Adventdalen measured OH transition lines between 8290Å and 8730Å (6-2 transition). The Half Meter Black Spectrometer in Ny Ålesund measured OH transition lines between 7250Å and 7450Å (8-3 transition). The One Meter Green Spectrometer was only used to double-check sky conditions.


For evaluating temperatures the background was determined and subtracted from the spectra. Then a model result of an OH-layer was fitted to the data with temperature as fit parameter. If necessary also aurora lines were added to the fit. The data was held trustworthy if the covariance of data and fit was larger than 0.900 and the background was modeled correctly.


For getting data of such good correspondence it is important to determine the sky conditions first. It is not possible to obtain results under conditions as strong aurora (large aurora lines in the spectra) or light pollution (high offset).


Spectrum Silver Bullet Spectrometer (Auroral Station)


                                    1    2    3    4    5    6     7    8       9  10                                           11

Figure 4.3.1: Spectrum from Silver Bullet Spectrometer


Spectral lines

line 1:               OH(6-2) Q                  8345Å

line 2:               1P                          8367Å             Aurora

line 3:               OH(6-2) P2(2)             8384Å

line 4:               OH(6-2) P1(2)             8401Å

line 5:               OH(6-2) P2(3)             8417Å

line 6:               OH(6-2) P1(3)             8433Å

line 7:               OH(6-2) P2(4)             8455Å             + OI 8446Å     Aurora

line 8:               OH(6-2) P1(4)             8467Å

line 9:               OH(6-2) P2(5)             8494Å

line 10:             OH(6-2) P1(5)             8506Å

line 11:                                           8630Å             Light pollution from Longyearbyen


for calculating temperatures lines 3 – 6, 8 – 10 have been used



Spectrum Half Meter Black Spectrometer (Ny Ålesund)























                                              1     2       3     4        5 6      7        8     9     10      11         12    13    14

Figure 4.3.2: Spectrum from Half Meter Black Spectrometer


Spectral lines

line 1:               OH(8-3) Q1(1)                        7276Å

line 2:               OH(8-3) Q1(2)                        7284Å

line 3:               OH(8-3) Q1(3)                        7293Å

line 4:               OH(8-3) P2(2)                         7303Å

line 5:               OH(8-3) P1(2)                         7316Å

line 6:               [OII]                                        7319Å             Aurora

line 7:               OH(8-3) P2(3)                         7330Å             Aurora

line 8:               OH(8-3) P1(3)                         7340Å

line 9:               1P(5,3)                               7349Å             Aurora

line 10:             OH(8-3) P2(4)                         7360Å

line 11:             OH(8-3) P1(4)                         7370Å

line 12:             1P(5,3)                               7384Å             Aurora

line 13:             OH(8-3) P2(5)                         7391Å

line 14:             OH(8-3) P1(5)                         7401Å


in addition there is a broad band from light pollution between line 7 and 14

for calculating temperatures lines 4 – 8, 10 – 11, 13 - 14 have been used


Different sky conditions and how they look like on the Silver Bullet Spectrometer


1. Cloudy, light pollution from Longyearbyen






¬ strong N line at 8630Å












Text Box:                     Figure 4.3.3: spectral quickplot













¬large offset: 35 counts/Å






                                                              ­ strong N line at 8630Å

Figure 4.3.4: Spectrum


2. Aurora and good conditions (no aurora)












¬ Aurora line 8446Å



¬ Aurora line 8367Å







                              ­                              ­

                      Nearly no aurora        Strong aurora

                      (spectrum 2)              (spectrum 1)


                      Figure 4.3.5: spectral quickplot



                               ­                        ­

                  Aurora line 8367Å       Aurora line 8446Å

Figure 4.3.6: spectrum 1 (strong aurora)



                              ­                          ­

                       Weak aurora line        Weak aurora + weak OH line

Figure 4.3.7: spectrum 2 (good conditions)


Temperature results from OH measurements with the Ny Ålesund spectrometer (1/2m black)



























































































































* these temperature measurements were excluded because the background is not modeled correctly


Data with a covariance below 0.9000 between measurement and fit is considered as not reliable. The calculated temperatures show a large variability, therefore average temperatures are calculated.





standard deviation




-- (only one meas.)


















all day




all day








Time dependency of temperature

The temperature data gives no indications for any time dependencies.

Figure 4.3.8: Temperature versus time for all data






























Figure 4.3.9: Temperature versus time for the first two days only


4.4. Lidar

Due to the limited data received from the LIDAR and spectrometer in NY-Ålesund, conclusions are not easily drawn. Neither the graph showing variations in temperature and height of the OH layer over time, nor the one showing temperature versus height, indicates any correlation between the two parameters (fig. 4.4.1 to 4.4.4 ).


Figure 4.4.1: Temperature versus height 5th November 2002


Figure 4.4.2: Temperature versus height 5th November 2002


Figure 4.4.3: Temperature versus height 6th November 2002


Figure 4.4.4: Temperature versus height 6th November 2002


In Fig. 4.4.3 linear dependence of increasing temperature can be observed from below 79 km. Because of so few measure points this might well be a coincidence, which is also strengthened by the low correlation in the other graphs, especially in Fig. 4.4.1 .


It appears to be no relation between the height and the temperature in the OH layer.

However, longer and more detailed data series might show different results. This part of the atmosphere is in no way stable and thus requires studying over a longer period of time.


4.5 Eiscat

The EISCAT-radar was used in order to get a reference temperature to the LIDAR measurements in Ny-Ålesund. The movable 32m-radar dish was aimed towards Ny-Ålesund at an angle which would allow the EISCAT radar to focus at about 90km height, the working range of the LIDAR. Even though using a suitable code and transmitting sequence the EISCAT radar proved incapable of obtaining reliable data from altitudes below 300km. Subsequently the EISCAT data could not be used for direct comparison with the LIDAR data. However the EISCAT data might be used as a cross-reference for temperature measurements at higher altitudes.


Figure 4.5.1: Electron density and ion temperature from Nov 13 2002-11-20, 32m

Figure 4.5.2: Electron density and ion temperature from Nov 14 2002-11-20, 42m



Figure 4.5.3: Electron density and ion temperature from Nov 14 2002-11-20, 32m


Figure 4.5.4: Ion temperature from Nov 13, series 43 of 173


Figure 4.5.5: Ion temperature from Nov 14, series 42 of 224


The electron density and ion temperature plots produced by the EISCAT radar (fig 1-3) show that it is very difficult to obtain reliable information about the ion temperature below 200-300km altitudes. To obtain more elaborate information the electron density in the lower parts of the atmosphere would have to be much greater, conditions which might occur during immense auroras and ion precipitation. At altitudes above 300 there is no problem getting reliable data something that might be used as a cross-reference to other instruments and atmospheric models.


Plotted on its own vs. altitudes below 300km the ion temperature (fig 4-5) acquired from the numerical EISCAT data present a highly incoherent source of information. The spreading of the error bars makes interpretation of the temperature very hard if not impossible. Examining all the scanning sequences from Nov13 and Nov14 these are the least noisy ones and for some reason presently unknown they are identical.

4.6. Meteor Radar

Entering earth’s atmosphere a meteor is heated up and evaporates. The produced hot and ionized gas is a good reflector for the radar signal of the meteor radar. The ionized gas will diffuse into the surrounding and the reflection signal disappears. From the rate of disappearing the diffusivity Da can be derived. The meteor radar is able to observe a height region from 80km to 105km.


The diffusivity Da changes in time and space. The diffusivity is low at the lower part of this height region about 2.5m2/s and increases with altitude to a value of ca. 12m2/s in a height of 97km. Above a height of 95km the diffusivity is approximately constant or even decreases, and the uncertainty of the measurements increases fast (see fig 4.6.1).


Figure 4.6.1 The diffusivity increases with height.


In a certain height diffusivity is varying with time. The variation in time is larger in higher altitudes (see fig 4.6.2). In order to see perturbations passing through the atmosphere the time series in fig 4.6.3 are normalized. The variations in 80 and 100km height are the highest. In 100km this might be also due to large uncertainties of measurements. On Nov 6th, 7th and 12th a maximum of diffusivity is observed in the late morning to noon in all layers, as well as on November 5th, 8th and 11th in all layers except that in 100km.



Figure 4.6.2 Diffusivity signal in the height of 80(black), 90(blue) and 100(red) km


Figure 4.6.3: Normalized diffusivity signal in the height of 80(black), 85(blue), 90(red), 95(magenta) and 100 (green) km


Fig. 4.6.4 shows a zoom in for the dates Nov 5th to 8th, 2002. On 5th and 8th the 100km (green) layer shows a reverse behavior than the other layers, however on 6th and 7th the changes in 100km are in phase with the other layers. The change of diffusivity may be caused be waves penetrating the middle atmosphere from earth’s surface to space. That results in diffusivity changes that first emerge in lower altitudes and move on to higher. For example, at the morning of 12th November the 80km layer shows a peak earlier than other altitudes. If the reason for diffusivity change is located in space then the change in diffusivity moves from higher altitudes to the bottom. At 6th, 7th and 11th of November the last layer showing the maximum is in 80km.


Figure 4.6.4: Normalized diffusivity signal in the height of 80(black), 85(blue), 90(red), 95(magenta) and 100 (green) km


The Fourier transformation of the diffusivity shows that a diurnal signal and a semidiurnal signal are observed from 80 to 90km. Also a signal with a period of 1.6 days is observed in 85 and 90km and a signal with 2-day variation in 100km altitude (see figure 4.6.5 to 4.6.9).


Figure 4.6.5: Power spectrum of the diffusivity signal in 80km altitude


Figure 4.6.6: Power spectrum of the diffusivity signal in 85km altitude

Figure 4.6.7: Power spectrum of the diffusivity signal in 90km altitude


Figure 4.6.8: Power spectrum of the diffusivity signal in 80km altitude


Figure 4.6.9: Power spectrum of the diffusivity signal in 80km altitude



In order to explain these changes a theoretical approach concerning the diffusivity in dependence of other physical properties is used (Hall).  Diffusion velocity, vd, is given by following expression:






,                                                                                         (4.6.2)


,                                                                                                             (4.6.3)


,                                                                  (4.6.4)


Te and Ti are the electron and ion temperatures, respectively, and P0 and T0 are the values of standard temperature and pressure.


As explained in the instrumentation chapter, temperatures in the meteor trail production altitudes can be derived from the diffusion velocity of the trails. Assuming the mobility parameter K0 to be constant (K0=2.5 x 10-4 m2/Vs), the derived temperatures are not reliable at lower altitudes, as illustrated by figure 4.6.10. A different approach is chosen here: the mobility of ions is studied by calculating how the mobility parameter must vary in order to produce more reasonable temperatures.


The altitude range for this study is chosen to be 77 km - 96 km , for following reasons:


1)      Temperature data seems to be reliable for altitudes from about 95 km upward.

2)      Below hundred kilometers we can safely assume the ion and electron temperatures to be the same, due to higher collision frequency.

3)      According to rocket measurements (figure 4.6.11) the polar winter temperature seems to be approximately constant at altitudes from 70 to 90 kilometers.

4)      Independent temperature values can be found from available OH-data, and the OH-layer height, although varying, falls in the chosen altitude region.

5)      Meteor radar data is available from 77 km upwards, and a range of 20 kilometers seems sufficient.




Figure 4.6.10: Temperature as a function of height, 13th November 2002.



Figure 4.6.11: Rocket measurements from von Zahn and Meyer, 1989.


After a constant temperature along with coinciding electron and ion temperatures is assumed, the temperature values are found from OH-layer measurements (see chapter 4.2 and 4.3). According to assumption the temperature might vary in time, but the mean temperature will not change with height. Temperature data is available for 5th and 6th November and obtained temperature values for OH-layer are 223 K and 209 K, respectively. Since a study of only two days seems insufficient, an average temperature value of 216 K is assumed when meteor data from other days is being plotted.


Figures 4.6.12 and 4.6.13 show the mobility parameter as a function of height for 5th and 6th November respectively. Error bars are added by taking into account the standard deviation in the temperature measurements (again the reader is referred to chapters 4.2 and 4.3). In figure 4.6.14 the plots are shown together without error bars.


At 96 km the mobility parameter has approximately the theoretical value of 2,5 x 10-4 m2/Vs, but down at 76 km it is already six times bigger. This shows that if the initial assumptions are correct and no major source of error in deriving the temperatures from meteor radar data is overlooked, the mobility of ions varies quite a lot in the region of interest, and when deriving the temperatures from meteor radar data, K0 should not be assumed constant. A reasonable explanation, why the mobility parameter should increase in lower altitudes, is missing.



Figure 4.6.12: Mobility parameter as a function of height, 5th November 2002.


Figure 4.6.13: Mobility parameter as a function of height, 6th November 2002

Figure 4.6.14: Mobility parameter as a function of height, 5th and 6th November 2002.


Next step is to investigate the change of diffusivity in one height level. The variation there may be due to a variation in mobility parameter or temperature.


When the mobility parameter is plotted as a function of time assuming constant mean temperature at a constant altitude, both diurnal and semi-diurnal variations can be seen. In figure 4.6.14 the mobility parameter is plotted at and altitude of 85 km for November 8th  (yellow), 9th  (black), 11th (green), 12th (red) and 13th  (blue). November 10th is left out because of some missing data. A temperature of 216 K is assumed for all days. Mean value seems to exhibit a fairly clear maximum at around local noon and approximately twelve hours later.


If the mobility parameter truly varied, its variation could in theory be due to the changes in ion composition in the atmosphere. Surely the ion composition changes with increasing and decreasing solar radiation, but solar radiation is scarce at high latitudes in winter. Also the varying amount of ionizing solar radiation could not explain the two maxima in one day presented in figure 4.6.14. Ion composition could also vary due to tidal waves in the atmosphere, which could explain the two maxima. Yet another possible source for variation would be the composition of incoming meteors exhibiting a daily pattern, but this speculation is highly theoretical.


If the mobility parameter is assumed to be constant at constant height, its apparent variation could in fact be produced by the changing temperature. Diurnal and semi-diurnal temperature variations due to tidal waves have been measured in wintertime at high latitudes in the upper atmosphere (Nielsen et al.). Magnitudes of these variations are 30 K and 15 K, respectively.

If the mobility parameter is now taken to be 5,9686 x 10-4 m2/Vs at an altitude of 85 km, a temperature change of approximately 69 K is required to produce the difference between the noon maximum and the following minimum. This is the of the same magnitude as the previously measured diurnal temperature variation.


One possible source of errors is the model atmosphere used in deriving the mobility parameter (equations 4.6.3 and 4.6.4). If the model is not correct for density, derived values of mobility parameter will be affected. It is also possible that the diffusion equation being used (4.6.2) is too simple to adequately describe the trail diffusing process at the region of interest.  




Figure 4.6.15: Mobility parameter as a function of time at 85 km, 8th-9th and 11th-13th November 2002.



5. Conclusions

The original goal which was to compare the data obtained from different instruments running simultaneously was only partly fulfilled because of difficult weather conditions and instrumental breakdowns. There were three days 5th, 6th  and 15th of November, 2002 when it was possible to calculate temperatures from the spectrometer measurements. The temperature varied irregularly with time. The average temperature was 216 ± 16 K.  The height of the OH layer was obtained from the lidar measurements for 5th and 6th of November but there was no correlation seen between the height of the OH layer and temperature. Temperature data from Eiscat could not be used in comparison with other instruments since the measurement region was above 100 km. It was not possible to get reliable temperature data from the meteor radar in the region of interest. Instead, temperature data from OH layer measurements was used to calculate the variation in the mobility parameter. It was apparent that the mobility parameter should not be considered as constant at the OH layer height. When the diffusivity coefficient was plotted as a function of time, both diurnal and semi-diurnal variations were seen. These variations may be caused by changes in mobility, density, temperature or meteor compositions.


One of the main purposes of the campaign for students was to get familiar with different kind of scientific methods and instruments when it comes to studying the upper middle atmosphere. During the running of Auroral Station students got a lot of practical training in operating and calibrating the instruments. Evaluation and representation of the obtained data was good practice for future scientific studies. Also very profitable for students was to see what kind of difficulties may occur while running the instruments and studying the results, and also to learn how to handle with these problems. Some preparation for potential future activities was also done by studying Mini-Dusty Rocket instrumentation, even though there was no possibility for launching the rocket during this campaign.



6. References


Brasseur, G., Solomon S., 1986. Aeronomy of the Middle Atmosphere. Second Edition, D. Reidel Publishing Company, Holland.


Hall, C.M., 2002. On the influence of neutral turbulence on ambipolar diffusivities deduced from meteor trail expansion. Annales Geophysicae (2002) 20: 1-6.


Nielsen, K.P., Deehr, C.S., Raustein, E., Gjessing, Y. and Sigernes, F., 2000. Polar OH-airglow temperature variations in the 87/88 winter. Phys. Chem. Earth, Vol. 25, No.2.


Olsen, S.V., Noteborn, D., Manual for preparing a Mini-Dusty payload.





Appendix A

“One Meter Green” Spectrometer – Cookbook / Instruction manual

General information


General procedures










Appendix B

“Silver Bullet” Spectrometer – Cookbook / Instruction manual

General information



General procedures









Appendix C

MSP (Meridian Scanning Photometers) – Cookbook / Instruction manual


Data:                           5 operational channels              

                                    Bandwidth 5 Å.

                                    Spectral resolution, typically 4 Å.

            Spatial resolution 1°.

                                    Spatial coverage 0-180° (N – S) along geomagnetic meridian.

                                    Scan time 16 s.

                                    Operating with rotating mirror tilted 45°

Mirror revolution time 4 s.



Channels:                    1. 6300 Å Atomic Oxygen (green color aurora).

                                    2. 4278 Å Molecular Nitrogen.

                                    3. 5577 Å Atomic Oxygen (red color aurora).                                                              

                                    4. 4861 Å Protons (Hydrogen Hb ).

                                    5. 8446 NIR (atomic Oxygen).


Start up:                      Secure the perimeter for polar bears.                 (Outside station)

                                    Put on slippers and make coffee.                                   (Station)

Check that the high voltage is off!

Enter the “sauna” and watch out for cables.                   (Sauna)           

Check that the mirror isn’t rotating.                              

Face the mirror upwards.                                             

Remove the covers from each filter, 5 individual.

Turn on the high voltage for all channels.                        (Station)

Carefully increase voltage to indicated level                              

on each channel.

Calibrate the MSP. See “Calibration”.                        (Station)

Start rotation of mirror. Make sure it starts.

If it doesn’t, press the black button to reset the            (Sauna)

fuse. Warning, do not use lights in the Sauna when the

high voltage is on!

Start the MSP measurements by typing “runmsp”.        (Station)



Calibration:                In DOS, type: “runcal”.                                                (Station)

                                    Let the filters calibrate 2-3 times. Press the “any” key to stop.

                                    Get the calibration coefficients for the filters.

                                    In channel 5, the aurora is the 1:st peak. The rest is OH.

                                    For channels 1 and 5, take the background value to the right.

                                    For channel 2-4, take the background to the left.

                                    Type: “config”.

                                    Input the calibration coefficients multiplied with 10,

also enter name and date.

Save by pressing “Alt+F”.


If you for some reason have to do the calibration once more,

you must first delete or move the previous calibration save-file.

The save-file is stored under “c:\unsaved”

and it is named by save year and day of year.

Ex: “cal02311.ly” ,  year 2002 day 311.

If you’re not familiar with DOS type:    

- cd..

- cd  unsaved

- dir

- move “filename” “new filename” or - del “filename”.

-cd \dap

- runcal



Shutdown:                   Type: “exit” to exit the program.                                    (Station)

                                    Decrease the voltage on all channels.

                                    Turn off the high voltage on all channels.           

                                    Turn off the mirror-rotating engine.                                (Sauna)

                                    Put on the five filter-covers. Don’t trip on the cables!



Appendix D

All Sky Video Camera (ALSC) – Cookbook / Instruction manual

The instrument is very sensitive for light, use it ONLY AT DARK TIME. Always close the door and turn off the lights. Close the hatch between the room and ALSC.


ALSC should turn on and off automatically. If not, the system can be started up by turning the red-coloured main switch, which is situated in the control panel on the left side of computer screen, and by opening the ALSC-monitoring program on computer. For shutting down the program first kill the ALSC-monitoring program and after that turn the main switch off.


Change tapes for recorder once a day. The second lowest recorder will start recording first and the lowest recorder after that. Before starting recording the time-counter for tape recorder should be set to zero by pushing the black button in the control panel for few seconds. Mark every recorded tape with place, season, date and the starting time of recording.


The real time all sky images can be watched from the monitors in computer room and kitchen.


The all sky images are also updated to Internet on a web page that can be found through the home-page of Auroral Station.


Appendix E - Timetable


Appendix F

OH-spectrums and fits (temperature calculation) – examples (one good and one bad fit)


[1] ALOMAR Observatory LIDARs.