Geodetic sensitivity to surface meteorological data: 24-h and 6-h observing sessions

In using any radiometric system, such as GPS, to determine terrestrial positions, it is necessary to account for the propagation delay of the neutral atmosphere (mainly the troposphere) if accuracies better than a few meters are needed. The zenith delay at sea level due to the hydrostatic (‘‘dry’’) component of the troposphere is about 2.3 m. By definition, this contribution can be determined from local surface measurements of barometric pressure to millimeter accuracy, limited by the measured refractivity constants for dry air. The standard model of Saastamoinen (1972) is normally used to compute dry delays given surface pressure data. At a given altitude, the relationship between pressure and dry delay is linear; at sea level, a 1 mbar pressure change corresponds to a delay change of about 2.3 mm. In the absence of local pressure data, global models for the dry delay can be reasonably accurate given latitude, ellipsoid height, and day of year; pressure varies over a range of up to about ±5%, leading to comparable errors in the modeled dry delay. The non-hydrostatic (‘‘wet’’) delay is much more spatially and temporally variable. For this reason, models for the zenith wet delay (ZWD), even given highly accurate surface met data, are notoriously unreliable. At sea level in temperate and tropical latitudes, the ZWD can reach 35 cm in summertime (i.e., 15% of the dry delay). In wintertime, the ZWD outside the tropics can sometimes approach 0, especially at higher latitudes. If uncalibrated, unmodeled, or inaccurately calibrated, tropospheric delays will induce an error in geodetic height measurements of a similar or larger J. Ray M. Morrison S. Hilla W. Dillinger G. Mader Geodetic sensitivity to surface meteorological data: 24-h and 6-h observing sessions