Solar Storms are Bad for Satellites, Right? MIT Researchers Not So Sure


PARIS — An analysis of 16 years of telemetry data from eight commercial telecommunications satellites in geostationary orbit has failed to establish a clear relationship between solar activity and failures of satellites’ power amplifiers, and concluded that a longer-term study of a larger population of satellites is needed to get to the bottom of the issue.

Massachusetts Institute of Technology (MIT) researchers, in what may be a first, persuaded a major commercial satellite fleet operator to turn over telemetry data on its satellites covering the 16-year period ending in 2012.

The eight satellites in question, representing two generations of spacecraft operated by Inmarsat of London, had reported 26 anomalies in their solid-state power amplifiers (SSPAs). In all but one of the cases, the SSPAs were sufficiently damaged to be switched off, and their functions were transferred to backup SSPAs.

Inmarsat, a provider of global mobile satellite services, provided the data after MIT agreed to legal restrictions on how it would be used. MIT has published the conclusions, “Space Weather Radiation Effects on Geostationary Satellite Solid-State Power Amplifiers,” in the current issue of the journal Space Weather.

Inmarsat has one of the most accident-free records of satellite launches and operations of any major operator. None of the 26 SSPA anomalies resulted in the spacecraft being taken out of service for any extended period. The eight satellites in question carried a total of 450 SSPAs.

While multiple studies have examined the impact of solar activity on satellite health, the MIT analysis may be the first using a full set of operator telemetry.

“What is more unique about our study is that we have partnered with [a] satellite operator to obtain telemetry, which is … very proprietary in nature,” Whitney Q. Lohmeyer, the report’s co-author, said Sept. 18 in an email. “Inmarsat was the first company willing to collaborate. With proper legal agreements and review processes, they have also enabled us to publish our work in conferences (both of scientific and engineering communities) and peer-reviewed journals.”

Sixteen years of daily telemetry data are a lot to comb through, but the paper’s authors say the limits inherent in a population sample of just eight satellites prevented them from drawing the conclusion many would have expected: that satellite power systems are damaged during periods of solar storms when plasma particles bombard spacecraft, depositing a charge on the satellite’s structure and causing electrostatic discharges and power drops or surges.

On multiple occasions over the years, satellite manufacturers have blamed coronal mass ejections for glitches on their satellites that resist other explanation. Satellite fleet owners have been at times skeptical about what they view as an easy out for builders seeking to avoid responsibility for workmanship shortfalls. 

The MIT researchers found that 16 years is not a long enough period when, as was the case with Inmarsat, not all the satellites under study have been in orbit for a full 11-year solar cycle.

In addition, five of the 26 anomalies could not be compared with the space weather environment at the time because of a lack of data. The principal sources of the weather data were NASA’s Goddard Space Flight Center Space Physics Data Facility’s OMNIWeb, an hourly compilation of solar wind, magnetic field and plasma data; and Los Alamos National Laboratory’s archive of electron-flux data.

The 21 SSPA anomalies that were successfully paired with space weather conditions at the time of their occurrence showed no obvious patterns that would invite easy conclusions. The satellites themselves were not identical, and were built by two different manufacturers, which may have added to the difficulty of the task.

The authors say the scientific consensus is that surface charging anomalies happen between midnight and dawn of the satellite’s local time. But seven of the Inmarsat SSPA anomalies occurred between noon and 5 p.m. local time.

Similarly, no correlation was found between periods when the satellites are in eclipse — between February and April, and between August and October each year — and the likelihood of an SSPA issue. “Interestingly, January was the month with the most anomalies, but is a time when geomagnetic activity is at a minimum,” the report said.

MIT is hopeful that publicity surrounding the study will encourage other fleet operators to contribute data. Lohmeyer said that two commercial operators, Telenor of Norway and Arabsat of Saudi Arabia, have already agreed to do so, and that a larger fleet operator has signaled its willingness to take part.

“This would not have been possible without Inmarsat’s willingness to engage in scientific and engineering research [and] to branch away from the traditions of proprietary telemetry and data management,” Lohmeyer said. “Their interest in improving their services and providing the means for manufacturers to improve their designs through working with researchers and students has and will enable steps to improving satellite communications. We are hopeful and confident that with other operators will follow in their footsteps.”