Characteristics of thermodynamic profiles derived from the GPS based Automated Meteorological Profiling System (AMPS)

The United States Air Force has acquired and is currently flight-testing the Automated Meteorological Profiling System (AMPS) at both Cape Canaveral Air Force Station in Florida and Vandenberg Air Force Base in California. AMPS is a GPS based radiosonde system that is designed to provide accurate, high-resolution profiles of wind and thermodynamic conditions at both major United States space launch ranges.

This paper describes the initial evaluation of the thermodynamic measurement capability of AMPS. The principal objective of this evaluation was to quantify the expected uncertainties associated with the operational measurements of pressure, temperature, and relative humidity. Since these parameters are reported as a function of height, the uncertainty in the height measurement of the AMPS radiosonde was also evaluated.

The evaluation strategy required three separate tests. First, the absolute uncertainty in AMPS temperature and relative humidity measurements was quantified in the manufacturers environmental chamber. Second, the relative uncertainty in AMPS temperature, relative humidity, and pressure measurements was quantified through a series of flight tests where two identical radiosondes were suspended beneath the same balloon so that simultaneous measurements by two instruments could be compared. Third, the absolute uncertainty in the AMPS height measurement was quantified through another series of flight tests were the height of the AMPS radiosonde was independently measured by two precision metric tracking radars. The results of all three tests were combined to provide an overall estimate of the thermodynamic measurement uncertainty of production representative radiosondes under representative operational conditions.

The test results have generally confirmed the expected performance of the AMPS radiosonde in terms of pressure, temperature, and relative humidity uncertainty. Temperature and relative humidity uncertainty is consistent with previous results for the Sippican small rod thermistor without radiation and lag correction and for the Sippican hygristor with correction. Pressure at each datum is estimated based on an integration of the hydrostatic equation. Since the radiosonde geometric altitude is based on the measured psuedoranges to at least four GPS satellites with differential corrections applied, the pressure measurement uncertainty is potentially better than most previous radiosonde systems and limited primarily by the absolute accuracy of the surface pressure at the release point.

However, an unexpected bias in radiosonde geometric height is observed. The GPS derived heights appear to overestimate the actual radiosonde height starting around 10 km and then increasing with altitude to as much as 150 m in some cases. The precise cause of this height bias is still unresolved. The current working hypothesis is the height bias is at least partially due to the method by which the differential corrections are applied to the GPS data. The reference receiver is located at the release point and surface based corrections are equally applied as the radiosonde ascends. This results in an increasing overcorrection of the tropospheric path delay at the higher altitudes and thus the observed altitude bias. However, the magnitude of the height bias seems to be larger than what would be expected from an overcorrection of the tropospheric path delay so other factors may also be contributing to this characteristic.