Start Date | Start Time | End Date | End Time | Data Quality Metric |
---|---|---|---|---|
04/11/1994 | 1432 | 05/31/1994 | 2359 | Suspect |
Subject: | SGP/SONDE/C1 - Sonde RESEARCH mode altitudes biased low |
DataStreams: | sgpsondewrprB1.a0, sgpsondewrprC1.a0, sgpsondewrprB5.a0, sgpsondewrprB5.a1, sgpsondewrprB4.a1, sgpsondewrprB1.a1, sgpsondewrprB4.a0, sgpsondewrprC1.a1 |
Description: | EDITOR'S NOTE: The C1 begin date is actually 4/7/94. This DQR applies to all sgpsondewXprFF.aN data, where X is r or n, FF is B1, B4, B5, C1, and N is 0 or 1. Discussion of Problem: I have discovered a problem with the RESEARCH mode (for PTU) Vaisala processing that results in a negative bias in the calculation of sonde altitude. This bias results from neglecting the sensed relative humidity when calculating air density when integrating the hydrostatic equation. In essence, the Vaisala RESEARCH mode (PTU) processing assumes a dry atmosphere when calculating sonde altitude. The magnitude of the bias is cumulative with height and will depend on the vertical distribution of moisture, but I have found that it can be as much as 20 meters at the tropopause (the RESEARCH-mode sounding altitudes will be smaller than altitudes calculated by using sensed RH). This problem applies to all soundings done in 'WXPR' mode, where X is either R or N. In particular, all soundings done during the April 1994 (RCS) IOP (4/7/94 to 5/21/94) are affected as well as those occasional inadvertent WXPR soundings. Note that the only variable affected is 'alt.' In response to my queries, Vaisala explained that the dry-air calculation was selected to increase the processing speed in RESEARCH mode. They consider the problem to be a minor one. Vaisala claims that altitudes in NORMAL (or NOMINAL) processing are calculated by including the sensed RH. |
Suggestions: | |
Measurements: | sgpsondewrprB5.a0: more
sgpsondewrprB5.a1: more sgpsondewrprB4.a0: more sgpsondewrprC1.a0: more sgpsondewrprB4.a1: more sgpsondewrprC1.a1: more sgpsondewrprB1.a0: more sgpsondewrprB1.a1: more |