A metal foil collection/flash desorption/flame photometric detection (MFC/FD/FPD) technique was used by investigators from the University of Idaho (UI) to measure ambient total sulfur gas concentrations from an aircraft platform during the NASA Global Tropospheric Experiment/Chemical Instrumentation Test and Evaluation 3 (GTE/CITE 3) program. The MFC/FD/FPD technique allowed rapid quantitation of tropospheric background air masses using sample integration times of 1--3 min with little or no gap between measurements. The rapid and continual sampling nature of this technique yielded data covering approximately 75% of the entire CITE 3 program's air track. Ambient air measurement data obtained during northern hemisphere (NH) flights often exhibited relatively high total sulfur gas values (up to 19 ppb) and an extremely high degree of sample heterogeneity, especially in coastal locations. Data from southern hemisphere (SH) flights typically exhibited relatively low total sulfur gas concentrations and a low degree of sample heterogeneity. A bimodal interhemispheric total sulfur gas gradient was observed using data obtained during transit flights between the two CITE 3 program ground bases. Comparisons were made of UI total sulfur gas measurements with composite sulfur gas values generated using speciated sulfur gas measurements from other CITE 3 participants. Only a relatively small number of overlap periods for comparison were obtained from all the available CITE 3 data because of large differences in measurement integration times and lack of synchronization of sample start/stop times for the various investigators.

These effects were compounded with extreme sample heterogeneity in the NH and the speed at which the aircraft traversed the air masses being sampled. Despite these constraints, sufficient overlapping data were available for the comparative evaluations. Comparison of NH UI total with composite sulfur gas values showed excellent correlation and linear curve fit, indicating substantial qualitative agreement. Simple linear regression of total on composite sulfur gas data yielded a slope of 1.9 for coastal NH regions and 1.2 for marine NH regions. The marine NH slope is not statistically different from one, indicating substantial quantitative agreement between UI total and composite sulfur gas values in these regions. However, a significant difference was observed when these same data were treated with a paired t test. SH data exhibited no significant correlation or linear regression slope. A paired t test showed a statistically significant difference when all SH flights were used. However, data from three SH flights that were classified into a unique group using discriminant analysis showed no significant difference between UI total and composite sulfur gas values when analyzed withthe paired t test. ¿ American Geophysical Union 1995