Atmospheric Effects
Every GNSS signal travels through two very different layers of atmosphere on its way to a receiver, and both of them bend and slow the signal in ways that RTK is built to cancel, but not always completely.
The ionosphere and the troposphere each delay a GNSS signal differently, and RTK cancels most of that delay by comparing a rover against a nearby reference. When conditions are unusually active or the baseline is long, that cancellation becomes less complete.
Two Layers, Two Kinds of Delay
The ionosphere, a layer of charged particles high above the Earth, delays GNSS signals by an amount that depends on solar activity and time of day, and that delay can vary noticeably over tens of kilometers during disturbed conditions. The troposphere, the ordinary weather-bearing layer close to the ground, delays signals based on local temperature, pressure, and humidity, and it can differ meaningfully between a base and a rover sitting at different elevations, even if they are otherwise close together.
When RTK Feels Different From One Day to the Next
A correction setup that performs beautifully on a calm, stable day can behave noticeably worse during a geomagnetic storm, or on a day with rapidly changing weather moving through the area. This is not equipment inconsistency. It is the atmosphere itself behaving less predictably, leaving RTK's differencing with a larger residual to contend with than usual.
What a Crew Can Actually Do About It
Atmospheric conditions cannot be controlled, but their impact can be managed. Keeping baselines shorter reduces exposure to atmospheric decorrelation. Avoiding especially low-elevation satellites, which travel through more atmosphere than one directly overhead, helps. Being aware of active space weather, published by several geomagnetic monitoring services, explains an otherwise mysterious bad day in the field before it gets blamed on the wrong cause.
A normally reliable RTK setup takes unusually long to reach FIX today, and briefly drops back to FLOAT twice during a routine survey. Is that necessarily a hardware problem? Not necessarily. Checking recent geomagnetic activity is a reasonable step before assuming the receiver itself has developed a fault.
The ionosphere and troposphere both delay GNSS signals, and RTK's precision depends on that delay being nearly identical for the base and the rover. Elevated solar activity, rapidly changing weather, and longer baselines all reduce how completely that delay cancels, producing real, external, day-to-day variation in RTK performance.
Atmosphere and geometry are conditions largely outside a crew's control. The next several chapters turn to something well within it: the physical setup of the rover itself.