Thousands of broadband satellites are quietly helping to estimate Earth’s upper atmosphere every single day, and nobody had to build anything new to make it happen. Tony Phillips, astronomer and operator of SpaceWeather.com, is treating the commercial satellite fleets already overhead as a distributed scientific instrument, reading publicly available tracking data to produce a daily index of thermospheric density. Solar activity is rising, and operators have already lost satellites to atmospheric swelling, which gives the project clear practical weight.
How the Method Works
Phillips converts publicly released orbital tracking files into a running atmospheric-density signal.
Even at low Earth orbit altitudes, the atmosphere is thin but real. Solar ultraviolet radiation heats the thermosphere, causing it to expand upward; as atmospheric density rises at satellite altitude, drag increases and satellites lose altitude faster. Phillips reads this signal through two-line element sets, or TLEs: the publicly released orbital files that describe catalogued objects in orbit for which suitable tracking data are available.
Each TLE contains a drag-related parameter called B*, which Phillips describes plainly: “B* is where the air density ends up.” When the real atmosphere is denser than the model assumes, the fitted B* value climbs to account for the extra resistance. Phillips converts those daily B* values into a sink-rate estimate through his own analytical procedure, calibrated against each satellite’s quiet-atmosphere baseline.
He analyzes roughly 1,000 Starlink satellites, based on a sample reported by The Register in September 2026, and cross-checks the results against Planet Labs SuperDoves, Amazon Kuiper satellites, and Eutelsat OneWeb satellites operating at different altitudes. As he explains on SpaceWeather: “No instrument was launched for this. The swarm itself is the sensor, read from public tracking data.”
What the Data Reveals
The satellite sink rate tracks solar activity with a measurable two-day lag, and geomagnetic storms produce sharper spikes still.
The sink rate follows the Sun’s 10.7-centimeter radio flux with an approximately two-day delay. Phillips describes that gap as the time the upper atmosphere needs to respond to changes in solar heating, though the precise lag reflects his specific analysis and is not a universal fixed constant. Geomagnetic storms, however, produce faster and sharper changes in drag.
The February 2022 event illustrates what that means in operational terms. SpaceX reported that drag rose up to 50% above prior launch baselines during a geomagnetic storm, and up to 40 of 49 newly deployed Starlink satellites were expected to re-enter or had already re-entered, according to SpaceX and CNBC reporting at the time. That episode is a clear real-world example of how thermospheric density changes at these altitudes are operationally consequential, not merely academically interesting.
What Is New and What Is Not
Phillips adds a public daily index where only single-event analyses existed before, though the underlying method carries real limitations.
Using satellite drag to study the thermosphere is not a new idea. The feature Phillips emphasizes is a continuously updated, publicly available index drawing on multiple commercial constellations at once, rather than retrospective analyses of individual storms. The method is indirect: it depends on orbit determination and model fitting, not a calibrated physical sensor.
Accuracy relies on the consistency of publicly available tracking data, as well as on orbit-determination quality, satellite attitude assumptions, and model limitations. The multi-constellation comparison is intended to help assess whether any observed pattern comes from one satellite design, one altitude band, or one tracking artifact, rather than from a genuine atmospheric signal. The available research does not establish that any satellite operator is currently using the index in their operations.
A Scale Advantage That Already Exists
The satellites are already there, already tracked, and already generating data on a regular basis.
The project’s core advantage is straightforward: the satellites are distributed across orbit and generate new tracking data repeatedly, without requiring any dedicated new hardware. If the index holds up over time, it could complement dedicated space-weather instruments and potentially inform how operators plan orbit maintenance and launch timing around periods of elevated atmospheric density. Phillips has also expressed criticism of some effects associated with large constellations, including astronomical interference and orbital congestion. Here, though, he is using the constellations as observational data sources, which says something interesting about how much science is already waiting in the data overhead.
