A previous version of this material was published in the refereed proceedings of the 19th Australian Space Research Conference (ASRC19) as:
Arie Verveer, David Coward and John Kennewell, "Unresolved Optical Observations of Material Degradation in Geosynchronous Satellites", Proceedings from the 19th Australian Space Research Conference, Adelaide 2019, pp47-54.


Even a large telescope is not able to resolve any detail of a object in geosynchronous orbit. Thus essentially all optical observations of geosynchronous satellites are limited to photometry, polarimetry and spectroscopy. Because of this there has been an effort in the last few years to try and characterise the nature of a space object in geosynchronous orbit from its light curve and its colour. The desired result of such observations is to try and determine the type, shape, size, origin and function of geosynchronous space objects. Related to this is the fact that the surface materials on orbiting objects are subject to degradation due to particulate and electromagnetic fluxes. We have undertaken low resolution spectral photometry - using standard astronomical filters - to follow the changes in spectral reflectance of a homogeneous group of geosynchronous satellites launched over a period of 20 years. We discuss the results and compare them to similar observations reported on a different set of satellites.


The orbital space environment has differences due to low gravity, vacuum, enhanced radiation and a plasma environment that affect materials in a different and sometimes more aggressive way than if the materials were at the surface of the Earth.

Enhanced radiation from galactic cosmic rays (GCR), solar particle events (SPE) and trapped radiation in the van Allen belts cause progressive reduction in output of solar cells as the total radiation dose increases. Much of this degradation is due to high energy particles penetrating the lattice of the crystalline silicon where they create lattice defects which reduce the quantum efficiency of conversion of solar photons into electrical energy.

This degradation may not be visible from the surface, at least for moderate doses. However, the vacuum of outer space allows access of the full range of solar ultraviolet, extreme ultraviolet and soft x-rays to the surface of the spacecraft, including the large area surfaces of the solar panels. Degradation due to this mechanism should result in noticeable changes in the surface properties of the spacecraft as a whole [1].

In low Earth orbit there are also some additional degradational processes due to the presence of a small residue of neutral atmosphere. These include sputtering and erosion due to atomic oxygen. Because spacecraft can be recovered from low Earth orbit, these effects have been extensively studied through direct examination of the affected surfaces [5].

No satellite has yet been retrieved from geosynchronous orbit, and the only possibility to study surface degradation due to the rather different space environment at these altitudes is through remote sensing. Because of the distances involved it is usually not possible to resolve a geosynchronous satellite in ground based imagery (exceptions are some recent interferometric studies [8]). Even large communication satellites subtend well under one arcsecond to a ground based telescope. Spectroscopy of the satellite as a point source appears to be the best means to examine degradational changes.

Spectral reflectance will be due to the surface properties of the satellite as a whole. However, as the largest proportion of the reflected light will come from the large solar panels, we are probably safe to assume that spectral differences between identically constructed satellites of different ages indicate differences in the optical surface properties of the two satellites, and that this difference is a product of space environmental effects.


The group of geosynchronous satellites chosen for this was the Chinese Feng-Yun 2 series of meteorological satellites (Figure 1). These spacecraft are spin-stabilized satellites rotating at 100 RPM with a de-spun antenna unit and body-mounted solar panels for power generation. Developed by the Shanghai Academy of Space Flight Technology (SAST) and China Academy of Space Technology (CAST), the cylindrical satellite bus measures about 2.1 meters in diameter and 4.5 meters in length. These satellites were chosen because they offer an homogeneous group of satellites of different ages for purpose of this study. The designation and launch dates of these satellites are shown in Table 1.

Feng-Yun 2 satellite
Figure 1: Generic illustration of the Feng-Yun 2 series meteorological satellite design.
[Image from National Satellite Meteorological Center - China Meteorological Administration]

At the time the imagery was taken FY2A and FY2C were below the observational horizon for several months. It is planned to image these at a later date. Only the last 3 satellites are currently active , the earlier ones having been placed in graveyard orbits as space debris. It should also be noted that FY-2H (not in the list) was launched last year after the initial study.

Satellite NORAD # COSPAR # Launch Date
FY2A 24834 1997-029A10-06-1997
FY2B 26382 2000-032A 25-06-2000
FY2C 28451 2004-042A 19-10-2004
FY2D 29640 2006-053A 08-12-2006
FY2E 35483 2008-066A 23-12-2008
FY2F 38049 2012-002A 13-01-2012
FY2G 40367 2014-090A 31-12-2014

Table 1. Designation and launch dates of the Feng-Yun 2 series of satellites


The observations were taken at the Skynet Meckering Observatory (Figures 2 & 3) located at latitude 31.638 degrees south and longitude 116.989 degrees east and elevation 197 m. This telescope is part of the University of North Carolina Skynet Network and is totally solar powered.

Skynet Meckering Dome Inside Dome
Figure 2: Skynet Meckering observatory external view showing the
solar arrays powering the observatory
Figure 3: 40cm RCOS telescope inside
the Skynet Observatory

The telescope used was an RCOS 40cm Ritchey-Chretien. The CCD camera is an Apogee U47 with 1024x1024x 6micron pixels. This gives a 10’ x 10’ image. Johnson series filters were used for low-resolution spectrometry[6]. The filters used were B(435nm),V(548nm), R(635nm) and I(880nm), where the figure in parentheses is the effective central wavelength of the filter. The filters were sequentially rotated into the light path for each satellite. For the newer satellites no tracking was required (as they were geostationary). For the older satellites which show considerable drift manual realignment was used after each exposure of 45 seconds. Photometry of the images was performed with Maxim DL software. All images were acquired in August 2018. Figure 4 shows one image of the Feng-Yun 2G satellite.

FY2 satellite image
Figure 4: Image of Fengyun 2G satellite taken with an infrared filter.
The satellite is the bright point near the centre of the image.


Table 2 shows the differential brightness (in magnitudes) of the various satellites as a function of their age in orbit. Error in magnitudes is +/- 0.02 mag. The differential brightness is the brightness of the satellite with the first colour filter minus the brightness with the second colour filter. Figure 5 plots this differential brightness with age for two filter combinations (blue/infrared and red/infrared). Only the change in differential magnitudes has any meaning in this context as apparent magnitudes are dependent on filter and telescope characteristics.

Satellite B - I B - V V - I V - R R - I
FY2B 2.30 0.87 1.44 0.72 0.72
FY2D 1.66 0.86 0.79 0.66 0.13
FY2E 1.46 0.85 0.62 0.49 0.13
FY2F 1.51 0.84 0.67 0.58 0.09
FY2G 1.57 0.84 0.72 0.67 0.05

Table 2: Differential Spectral Magnitudes for the 5 Geosats Measured

Results graph
Figure 5: Satellite differential brightness as a function of age.
Blue and red magnitudes are compared to the infrared magnitude.


The largest magnitude change with age occurs between the red and the infrared. This difference appears to be stable for the first decade of life but then appears to increase with the age of the satellite. On the other hand the difference between the blue and visible filters appears to be very stable. The difference between visible and red shows a smaller increase with age, but smaller than that between red and infrared. This would seem to suggest that most of the surface changes occur at the lower energy levels. Many spectral reflectance studies have been made on geosynchronous satellites, but many of these have been confined to the infrared [2].

It is interesting to compare these results with a similar study by Schmitt [3] (figure 6) made on four different groups of satellites.

Schmitt study
Figure 6: Graph from Schmitt (2018) showing spectral brightness differences for
four groups of satellites manufactured by different companies.

These either show a more random variation or else (in the case of the Loral satellites) an opposite variation. All of these satellite groups however, are not as homogeneous as are the FY 2 series. Other reasons for the smaller spread in our study might be that all our measurements were taken within a short period of time, with a consequent smaller spread in solar phase angle. It has been shown that relative spectral magnitudes can show a considerable variation with this angle [4]. Because the measurements for a single satellite were taken within a few minutes, we can discount any variations due to change in phase angle, but this is not so for the older satellites which are not controlled and may exhibit attitude changes. However, in the latter case we may expect the results to be reasonable because of two factors. All satellites were initially spin-stabilised at 100 rpm. This means that for an exposure of 45 seconds we are really measuring an average brightness over quite a substantial surface area of the satellite. This will partially negate any phase angle change due to an attitude change. Secondly, we expect the satellite to still have retained a reasonable fraction of its initial spin even over 20 years - maybe as much as 50 rpm. It is possible that additional rapid motion was given to the older satellites when they were kicked up to higher orbits. We can see this is the early Aussat communication satellites, but we see no evidence of this in the two oldest FY2 satellites which show large drift rates.

However, a much larger sample of measurements over a wide variation of phase angles will be necessary before we can claim that the observed variation with age is truly due to space environmental degradation of the spacecraft surface materials over the satellite bus lifetime. This is planned.


We would like to acknowledge Dan Reichart and the Skynet Telescope Network which made these observations possible. We would also like to acknowledge Ariane Group for suggesting the Feng-Yun 2 series of satellites for this spectral study.


  1. Jacqueline A Reyes & Darren Cone, "Characterization of Spacecraft Materials Using Reflectance Spectroscopy", Proceedings of the AMOS Technical Conference 2018, < >
  2. James Frith, Phillip Anz-Meador, Heather Cowardin, Brent Buckalew & Susan Lederer, "Near-Infrared Color vs Launch Date; An Analysis of 20 Years of Space Weathering effects on the Boeing 376 Spacecraft", Proceedings of the AMOS Technical Conference 2015, <>
  3. Henrique R Schmitt. “Multi-band optical photometry of geosynchronous satellites”, Proc SPIE 10641, Sensors and Systems for Space Applications XI, 1064104 (2 May 2018), doi:10.1117/12.2305276
  4. Andrew Jolley, Gregg Wade & Donald Beddard, "Multi-colour Optical Photometry of Active Geostationary Satellites", Proceedings of the AMOS Technical Conference 2015, <>
  5. D Hastings and H Garrett, Spacecraft-Environment Interactions, Cambridge University Press, 1996.
  6. Information on the Johnson-Cousins spectral filters used can be found at <>
  7. Matej Zigo, Jiri Silha and Stanislav Krajcovic, " BVRI Photometry of Space Debris Objects at the Astronomical and Geophysical Observatory in Modra", Proceedings of the AMOS Technical Conference 2019, <>
  8. Zachary J DeSantis, Sam T Thurman, Troy T Hix and Chad E Ogden, "Image Reconstruction from Data Collected with an Imaging Interferometer", Proceedings of the AMOS Technical Conference 2017, <>

Australian Space Academy