Results from the balloon ozone intercomparison campaign (BOIC)

Abstract

The purpose of the Balloon Ozone Intercomparison Campaign (BOIC) was to assess the accuracy and precision of various ozone measurement systems under development and those flown operationally. Ozone observations made by in situ UV absorption photometers from four groups, two solar UV absorption photometers, three varieties of electrochemical sondes, and a mass spectrometer were intercompared in three flight missions, each involving several balloon flights. Concurrent Umkehr and satellite observations were also intercompared. The National Bureau of Standards (NBS)provided a reference ozone source for intercomparing several of the in situ instruments at ground pressure. Harvard University and the University of Minnesota each developed a calibration facility after the BOIC flights which enabled a comparison of the flight in situ photometers under simulated stratospheric pressure and ozone concentrations. The following is a summary of results. The standard deviation of the sensitivities among 18 instruments tested about the NBS reference was 11%. These differences appear in flight at the lower altitudes, but they change at higher altitudes, indicating height dependent errors, particularly for the electrochemical sondes. Comparisons at the two laboratory facilities simulating stratospheric conditions indicated differences of about 2% at 3 mbar for two of the flight in situ photometers. Ozone values among four in situ UV photometers flown together had a standard deviation about the mean value of about ±3% from the tropopause to about 41 km. This is very close to the expected accuracy of these measurements and therefore is an excellent result. However, during float at 42 km the difference nearly doubled. Experiments performed during flight demonstrated that even near 40 km, wall losses in three UV photometers were less than 5%, which is consistent with the laboratory simulations. A comparison of a complement of electrochemical sondes yielded results that agreed with the mean of the in situ UV photometers to within 0–20%, depending on the sonde type and altitude. The electrochemical sondes gave systematically lower ozone values at pressures lower than 10 mbar (above 31 km). A comparison between an in situ UV and a solar absorption photometer indicated a 10% difference in the stratosphere, where the in situ measurement was lower. Comparisons of electrochemical sondes and Umkehr and Solar backscattered ultraviolet (SBUV) satellite observations showed agreement to within their error bars (stated to be about 5–10%). In two comparisons between in situ UV photometers and satellite measurements, a consistent difference occurred between 10 and 3 mbar (28–40 km), implying a possible bias between the measurement types. Intercomparisons among all the instruments in the troposphere showed 20–30% differences from the mean. A comparison of pressure measurements performed by several experimenters resulted in differences as high as ±15% from the average measurement.

Ozone measurements from the Balloon Intercomparison Campaign

“…The NASA JSC UV photometer used in BIC was compared with five other UV photometers during the Balloon Ozone Intercomparison Campaign (BOIC) [Hilsenrath, et al (1985); Hilsenrath et al (1986)]. …”

“…Random and systematic errors (95% confidence level) in the cross section at 254 nm are estimated to be 6% (R. D. Hudson, personal communication, 1986). Mauersberger et al (1986) recently measured the cross section of ozone at 254 nm with a 95% confidence level uncertainty of 1.2%. The more conservative uncertainty of Hudson is used here.…”

“…Using the cross section uncertainty estimated by Hudson, the total 95% confidence level RSS uncertainty is 6.4-7.3% between 22 km and 40 km. If the Mauersberger et al (1986) cross section and uncertainty had been used then total RSS uncertainties would be less than 3.5% without consideration of losses. The BIC-I UV photometer profile is shown in Figure 2.…”

Study on Comparison of Indian Ozonesonde Data with Satellite Data

“…This is in contrast to the Brewer-Milford type of electrochemical ozone sensor [13]. Various studies using a standard UV photometer have shown that concentrations of tropospheric ozone are overestimated with ECC sondes, but underestimated by Brewer Mast Sondes [16][17][18][19][20]. The two error sources mainly affecting the tropospheric part of the profile are the contamination of the sonde sensor with reducing agents (SO 2 and NO x ), causing a negative bias especially for Brewer Mast sondes, and the estimation of the background current, causing presumably a positive bias for ECC sondes and a negative bias for Brewer Mast sondes [21].…”

“…Applying the full background current correction leads to an underestimation in stratospheric ozone of amount 2-6 % [9], possibly indicating a need to separate corrections for tropospheric and stratospheric data. Intercomparisons in 1970Intercomparisons in , 1978Intercomparisons in and 1984 had shown that ECC sondes measured more ozone in the mid-troposphere than BM sondes by about 15-20 % (with a range of 7-8 %); in the stratosphere, BM sondes gave more ozone than ECC sondes, usually by less than 5 % [4,5,17]. An intercomparison in 1991 gave opposite results [8].…”