In GNSS terminology, observations refer to the raw measurements that a receiver extracts from satellite signals, the fundamental data upon which all positioning calculations depend. These measurements capture the physical properties of radio waves traveling from satellites to the receiver antenna, providing the distance and velocity information needed to compute position solutions ranging from meter-level navigation to centimeter-level surveying accuracy.
The three primary GNSS observation types serve different positioning purposes. Pseudorange observations measure the apparent distance to each satellite based on the travel time of the signal’s code modulation, providing meter-level ranging precision suitable for standard navigation. Carrier phase observations track the accumulated phase of the carrier wave itself, achieving millimeter-level ranging precision essential for RTK and other high-accuracy techniques, though with an inherent integer ambiguity that must be resolved. Doppler observations measure the frequency shift of the carrier wave caused by relative motion between satellite and receiver, providing velocity information and aiding in carrier phase tracking.
Observations are collected at regular time intervals (epochs) and recorded with precise timestamps synchronized to GNSS time. Professional receivers may output observations at rates from 1 Hz to 100 Hz depending on application requirements. Each observation includes metadata such as the signal type (L1, L2, L5, etc.), satellite identifier, signal quality indicators (carrier-to-noise ratio), and tracking status flags. The quality and completeness of observation data directly determines achievable positioning performance.
For precision positioning applications, observations from multiple sources are combined through differential processing. RTK systems difference observations between base station and rover to cancel common errors. Network RTK uses observations from multiple reference stations to model error variations across a region. Post-processing workflows combine rover observations with reference station data for offline position determination. In all cases, the quality of raw observations, affected by factors including signal strength, multipath, interference, and atmospheric conditions, ultimately limits the accuracy of derived positions.
