IRIDIUM FLARES

Observational Project

 

 

Kaarin Goodburn

Registration No. 99616541

 

 

 

Figure 1: Photograph taken by Chris Dorreman

20 Sept 1997 at 19:10:23 UT. Believed to be one of the earliest pictures taken of an unpredicted

 (before any Iridium flare prediction programs were published) 

Mag -8 (estimated) flare produced by Iridium 12 (NORAD 24837/COSPAR 97-030-B).

IRIDIUM FLARES

Observational Project

1.       Introduction

 

This project describes the background to, mechanism for and accuracy of predictions of Iridium flares.

1.1     What are Iridium Flares?

 

Iridium satellites have a normal brightness of +6 magnitude, and are therefore at the limit of detection with the unaided human eye, but occasionally some of them produce reflective flares up to magnitude –8 (Figure 2), caused by one of the satellite’s mirrored Main Mission Antennae catching sunlight and reflecting it down to the viewer.  The mechanism of flare production is described more fully in section 1.4.

 

Venus can be as bright as magnitude -4.4, therefore Iridium flares can be nearly 30 times brighter than Venus, i.e. 2.512^(-8+4.4). Comparative magnitudes for a number of objects are given in Table 1 below.

 

The flares can last from 5 to 20 seconds before the satellite once again becomes almost invisible to the naked eye. Some flares have been observed during daylight (twilight) hours, which is very unusual for reflective glares from satellites.

 

However, the satellites cause some problems for radio astronomers who picking up interference from them.  This problem is exacerbated since the satellite orbits are designed to give coverage of the entire Earth, therefore affecting all radio telescopes everywhere on Earth. 

Figure 2: Animation of Iridium Flare with Meteor

 

1.2     Magnitude

 

Magnitude is a measure of the brightness of a celestial object. The lower the value, the brighter the object, so magnitude -4 is brighter than magnitude 0, which is in turn brighter than magnitude +4. The scale is logarithmic, and a difference of 5 magnitudes means a brightness difference of exactly 100 times. A difference of one magnitude corresponds to a brightness difference of 2.512 (the fifth root of 100).

 

The system was started by the ancient Greeks, who divided the stars into one of six magnitude groups with stars of the first magnitude being the first ones to be visible after sunset. In modern times, the scale has been extended in both directions and more strictly defined.

 

Table 1: Apparent Magnitudes

 

 

 

 

1.3     What are Iridium Satellites

 

The Iridium satellite constellation was a key part of Motorola’s $5-billion project to create a worldwide mobile telephone network.  Motorola’s plan was to place 77 satellites in Earth orbit in seven evenly spaced orbital planes, with 11 spacecraft in each. Hence the name Iridium, whose atomic number is 77. Seventy-nine iridium satellites now orbit the Earth, including spares.

 

The three-sided Iridium satellites are relatively small, being approximately 4 metres long and less than one metre wide.  The axis of the satellite body is maintained vertical to the Earth's surface, which, together with the maintenance of its longitudinal position, allows the flares to be predicted.

 

The first five Iridium satellites were launched on May 5, 1997, aboard a Delta II rocket. All five satellites were spotted in close formation by observers within a short period after the launch (see Figure 3).

Figure 3: Iridium 22, 23, 24, 25 and 26

 

 

Figure: August 26, 1997, from Houston, Texas: Five Iridium spacecraft (from left to

right, numbers 22 through 26) while crossing the southernmost reaches of Cetus.

90-second exposure at 11:02 UT. Shown here shortly after launch, the spacecraft

have since separated to occupy separate positions in a common orbital plane.

Source: Sky and Telescope 1998

 

 

The satellites were initially placed in a circular orbit at an altitude of approximately 500 km. Over a two week period they were individually raised to a circular orbit of approximately 780 km by their own onboard propulsion systems.  In its 100-minute operational orbit, each satellite circles the Earth 14.34 times a day.

 

The expected lifetime of an Iridium satellite is 5-8 years.

 

 

1.4     Flare Mechanism

 

The mechanism providing the flare/glint is the three Main Mission Antennae (MMA) on each of the satellites, which were built to relay digital traffic at frequencies from 1616 to 1626.5 megahertz.

 

These antennae, which are 120° apart on the satellite and are 188 cm wide x 86 cm long x 4 cm thick, are highly reflective aluminum flat plates treated with silver-coated teflon for thermal control that are angled 40° away from the axis of the body of the satellite (Figure 4).

 

The plate or MMA can act as mirrors providing a direct (specular) reflection of the sun's disk, which is only tens of kilometers wide at the Earth's surface. Therefore, for an observer to see a very bright flare, they must be within this relatively small area. Prediction programs are available to determine this area.

 

Figure 4: Iridium Satellite

		Another possible source of flares off the Iridium satellites has been referred to by 

		Alan Rohwer, which may account for unexplained reflections not predicted by the 

		prediction programmes.

		Rohwer reported on SeeSat-L (23/10/97) that two other surfaces on the satellites 

		are coated with flexible optical solar reflector like the MMAs. Both surfaces are 

		located on the main bus and on the plus X: 

·         The equipment panel (86 x 188 cm) underneath the MMAs  - quite well shielded 

	by the MMAs as viewed from the ground, and 

·         The Battery Radiator Assembly (BRA), which is about 78 x 84 cm. The BRA is 

	located on the upper bus between the solar arrays and may catch and reflect 

	light under some illumination angles.

Both of the surfaces are located on lines that are along the earth radius line. 

 

2.       Motivation & Aims of this Project.

 

The potential controlled deorbiting of the Iridium satellites owing to the bankruptcy of Iridium LLC and the concomitant loss of the flares was the main driver behind my interest in undertaking this topic, coupled with the sheer spectacle of seeing a major flare.

 

The main aim of the project was to assess the accuracy of the predicted flares. A secondary aim is to provide information regarding the current state of some of the satellites.

 

3.       Method Used

 

There are a number of prediction programmes  available to predict Iridium flares. I used the Heavens-Above website, which is based in Munich, Germany, describing my location as ‘London’.

 

The programme provided my co-ordinates as 51.570°N, 0.1050°W.

 

On 5 November, observations were made from Birmingham, for which Heavens Above gave the co-ordinates as 52.485°N, 1.86°W.

 

On 13 and 14 November, observations were made from Wansford, Cambridgeshire, for which Heavens Above gave the co-ordinates as 52.582°N, 0.413°W.

 

Observing and recording equipment comprised:-

 

·         Digital stopwatch to record time of first sighting of the satellite (and/or flare, depending on which was visible first) – synchronised with the speaking clock before each observation

·         Compass

·         Notepad and pen

·         Initial electronic observation log (Outlook 2000 task function)

·         SkyGlobe programme

·         Final data log (Excel 2000 spreadsheet)

 

The expected location of the flare in the sky, as predicted by Heavens Above, was determined before the flare was expected, and that location observed closely for some minutes beforehand in order to acclimatise vision.

 

The time that the satellite was first spotted was noted, together with

 

·         Viewing conditions

·         The duration of the flare

·         Its maximum brightness

·         Whether it was a sudden burst of light or gradually built/faded, and

·         The approximate path (direction and length in degrees of arc) of the flare in relation to visible stars

 

The path of the flare in the sky was drawn on a SkyGlobe-generated printout of the area of sky in question, which was also used to pinpoint the predicted point of appearance of the flare.

 

Apparent magnitude was assessed by visual comparison with known stars and/or planets also visible at the time (see Table 2). Although this is inherently imprecise, it was found to give a reasonable measure of the accuracy of the flares’ predicted magnitudes.

Table 2: Source Data: Apparent Magnitude Comparisons

 

 

Sources: Venus and Jupiter - Astronomy Now; Others - Philip’s Astronomy Dictionary

 

4.                 Raw Data

 

A copy of my observation log is given in Table 3 below. Photocopies of tracks of each sighting are enclosed.

 

Table 3: Observation Log 10 September-26 November 2000

 

 

5.       Analysis & Interpretation

 

Poor weather conditions thwarted 21 out of 46 (>45%) of planned observations. However, the data from the remaining 25 observations did show a number of interesting features:-

 

 

1.                  The predicted appearance time of flares was less accurate than I expected:

 

14/25 = 56% on time

 

2.                  The predicted magnitude of flares was generally accurate:

 

18/25 = 72% predicted accurately

1/25 = 4% brighter than predicted (I 19, observation #38)

1/25 = 4% dimmer than predicted (I 29, observation #13)

 

3.                  A significant number (8/25 or 32%) of flares gradually grew in magnitude, i.e.

 

·         I 12, observation # 41

·         I 13, observation #42

·         I 19, observation #38

·         I 35, observation #33

·         I 36, observation #39

·         I 57, observation #29

·         I 61, observation #36

·         I 86, observation #43

 

The remainder were not visible prior to the main flare.

 

4.                  I 30 (observation #27) showed a double flare. The first was 11 secs earlier than the prediction, and the second one second later.

 

5.                  Some flares were so slow-moving that they appeared to be almost stationary, e.g.

 

·         I 35, observation #33

·         I 19, observation #37

 

6.                  Flares generally tapered off in magnitude:

 

·         I 22, observation #3

·         I 49, observation #8

·         I 35, observation #33

 

7.                  However, some disappeared suddenly, e.g.

 

·         I 28, observation #32

 

8.                  Flares could last for varying periods of time, e.g. range of

 

·         1 sec for I 30 (observation #27)

·         40 secs for I 32 (observation #12)

9.       The length of visibility of satellites varied widely, although for post-sunset observations limiting visibility was generally the same (~mag 4.5).

 

·         The range of periods of satellite visibility: 4-55 secs

 

 

9.                  The apparent speed of satellites varied (apparent travel in degrees of arc/duration of visibility), although each satellite observed is at the same height above the Earth and orbits at the same speed.

 

 

10.             Flares did not always appear as predicted, despite apparent good viewing conditions. For example:-

 

·         I 28, observation 12: mag 0 predicted, but not seen – possibly since dawn sky was too light already

·         I 31, observation #21: mag –1 predicted, but not seen – ditto

 

This would bring into question the criteria used by Heavens Above for determining what objects would be visible in dawn/dusk skies, beyond any weather and/or pollution effects.

 

 

6.       Conclusions

 

My results bear out in broad terms with reports given on the SeeSat-L mailing list, particularly in terms of the double flare seen (observation #27) and accuracy of predictions in terms of timing. These observations can be accounted for, respectively, by

 

1.       Flares/glints off parts of the satellites other than the MMAs (see section 1.4), and

2.       Anomalous orbits of some satellites, affecting predictions

 

The observations carried out were reliable in terms of timing accuracy to within a second, but the apparent travel of satellites across the sky in terms of degrees of arc was difficult to assess in absolute terms. However, this aspect of the observations was almost incidental as the accuracy of the predicted timing, magnitude and initial location of the flares were the focal points of my work.

If I were to re-do the observational work I would:

 

a)       Compare predictions made by different programmes

b)       Compare the effect of specifying more precisely my location rather than stating ‘London’ when using a programme

c)       Involve a second person to start a stopwatch when the flare, rather than the satellite, first became visible, and therefore time more precisely the duration of the flare.

d)       Devise an accurate method of determining degrees of arc traveled by the satellites (and/or flares) when visible.

e)       Use a video camera to record the observations, for later playback and analysis

f)        Select a more reliable observing location in terms of weather!

 

 

7.       End Note: The Future of the Flares

         On 15 November 2000, the US Bankruptcy Court for the Southern District of New York 

         approved the bid of Iridium Satellite LLC to purchase the operating assets of Iridium LLC 

         and its subsidiaries, including the Iridium satellite constellation, the terrestrial network, 

         Iridium real property and intellectual property owned by Iridium LLC.  The company was 

         reported at that time to have contracted with the Boeing Company to operate and maintain 

         the satellite constellation. 

 

It therefore appears, at the time of writing, that Iridium flares will continue, although it is unclear what measures will be taken by Iridium Satellite LLC to repair or replace those satellites known to be either tumbling or malfunctioning in some other way.  It is possible that since the satellites were originally launched to provide ‘whole Earth’ coverage, further satellites will need to be orbited to replace those currently not functioning (Table 3) and those that expire in the future.

 

Table 3: Iridium Satellites Currently Known to be Non-functional

 

8.       References & Further Reading

 

Astronomy Now     ISSN 0951-9726

 

Ed Cannon             SeeSat-L: Iridium 21 (97-34E, #24873) -- NO FLARE!

 

Iridium Satellite LLC Press Release, 15 November 2000

 

Iridium LLC            Press Release, 3 March 2000, Washington Post 3 March 2000

 

Magnitudes            www.satellite.eu.org/sat/vsohp/magnitude.html

 

Alan Rohwer          SeeSat-L: Additional information on Iridium satellite construction Additional information

 

Alan Rohwer          SeeSat-L: Warning!  Some of the current Iridiums apparently are not reliable flare/glint/flash producers. www.satellite.eu.org/sat/seesat/Oct-1997/0349.html

                  

John Woodruff (Ed.) Philip’s Astronomy Dictionary, ISBN 0-540-07759-5