Comparison of Two Radiosonde Pressure Measurements: Pressure Sensor vs. GPS-derived Pressure

Radiosonde measurements of atmospheric pressure are important both as the vertical coordinate for temperature, humidity, and wind profiles from the radiosonde, and for understanding the evolution of weather systems. Radiosondes use two main principles for determining the atmospheric pressure: one is a direct pressure measurement with a sensor; the other derives pressure from the radiosonde measurements of GPS height, temperature, and humidity through the hypsometric equation. While the two methods are accepted for operational use and they have been shown to agree within a specified measurement uncertainty in various test campaigns, the differences in their physical assumptions and accuracy may have an impact in some conditions. For example, the GPS-based method assumes hydrostatic balance, while the performance of the pressure sensor is limited by the required large dynamic range. In this study we describe the characteristics of each measurement method with results from a theoretical analysis and test soundings, and discuss the best method selection for various radiosonde applications.

The calculations show that a persistent temperature bias of 0.1 – 0.2 Celsius causes an error of up to 0.15 – 0.3 hPa in the GPS-derived pressure profile. A short-term temperature error due to wet-bulb conditions when the radiosonde emerges from a cloud produces an error of similar magnitude. The two methods may show differences in highly non-hydrostatic conditions, as demonstrated by soundings during tropical cyclones in Taiwan in 2014.

Results from the test campaigns in Malaysia and Finland are presented using Vaisala Radiosonde RS41, comparing radiosonde configurations with a pressure sensor against radiosonde configurations with GPS-derived pressure. The average differences for sensor and GPS-based pressure measurements are less than 0.2 hPa. The results indicate a good overall agreement between the two methods, while the presented analysis suggests the preferred method for specific goals.