This thesis consists of three distinct parts of CAST, CONTRAST and ATTREX, which were aircraft and field campaigns in the West Pacific in January-March 2014. The first section comprises of ozonesonde measurements from Manus Island, Papua New Guinea. A contamination issue affected the first 14 ozonesondes, and so particular care was required to characterize the background current, and as a result, a 'hybrid' background current correction was developed, which combines a constant correction with a pressure dependent correction. Collocated measurements with the CONTRAST aircraft - the NCAR Gulfsteam V - suggests the new hybrid correction produces better ozonesonde profiles than the other corrections that are found in the literature. The results of the ozonesonde measurements revealed a low-ozone event, with minimum ozone concentrations of ~12 ppbv, which was coincident with an easterly jet, and traced back to an area of deep convection: clean marine boundary layer air was uplifted into the tropical tropopause layer (TTL) and then advected in the easterly jet across to Manus Island. The second section attempted to find more examples of low-ozone conditions in the TTL from the aircraft data. The ATTREX aircraft - the NASA Northrop Grumman Global Hawk - observed ozone concentrations of ~10 ppbv in the Southern Hemisphere in proximity of tropical storm Lusi. Whole air samples from all three aircraft suggests the low-ozone air had recently encountered the boundary layer, with enhanced concentrations of surface-generated very short lived substances (VSLSs) compared to air with higher ozone concentrations. No low-ozone events were found in the Northern Hemisphere, even in the vicinity of tropical cyclone Faxai. The third section explores the low-ozone events in the WRF-Chem (Weather Research and Forecasting - with chemistry) in order to see whether the model was capable of recreating the low-ozone event measured by the ozonesondes on 21-23 February as a case study. The WRF-Chem simulation did correctly reproduce the large convective storm in a similar area to that observed by satellites, and surface tracers were uplifted in large quantities as hypothesized. However, no evidence of injection of air into the stratosphere was found in the simulation, and, rather than uplift directly from the surface, mixing of air in the boundary layer followed by uplift into the TTL was the main mechanism for producing the low-ozone event.