Examining the Relationships between Ozone, Water Vapor, and Vertical Velocity in the Tropical Western Pacific during the CONTRAST Campaign and Their Representation in MERRA-2 Reanalysis

Many regions of the tropics experience intense deep convection, including the region known as the Tropical Western Pacific (TWP). Convective heights of 10 km to 15 km above sea level have been observed in this region (Pan et al.2015). These convective heights have been attributed to persistent deep convection, leading to convective transport of pollutants from the surface to the mid- and upper- troposphere (Fueglistaler et al.1999, Panet al.2015). Convectively transported air has been associated with low water vapor mixing ratios and previous studies have shown a negative correlation between ozone concentration and water vapor mixing ratio (Pan et al.2015, Anderson et al.2016).

The 2014 NSF Convective Transport of Active Species in the Tropics (CONTRAST) research flight campaign investigated the impacts that both deep convection and convective transport have on the chemical composition of the TWP (Pan et al.2017). The in-situ data from CONTRAST provides a wide range of measurements, including ozone, chemical tracers, water vapor, and vertical velocity. Using the data from the CONTRAST campaign, this study will examine the relationship between ozone, relative humidity (RH), cloud presence, and vertical velocity. This study will also compare the in-situ results with reanalysis data from the Modern-era Retrospective analysis and Research Application, version 2 (MERRA-2).

Previous analysis by Fueglistaler et al.(2009) found an area of enhanced in-situ ozone concentrations between 150 hPa and 200 hPa in the tropical Pacific. Based on these results, we separated the ozone data from the CONTRAST campaign based on cloud presence and found the enhanced ozone concentrations (above 100 ppbv) are only observed for “clear-sky” conditions, while the ozone concentrations for “in-cloud” conditions did not exceed 100 ppbv. Further examination of in-situ ozone was done by creating a frequency plot that shows a bimodal distribution, with a primary mode located at 20 ppbv and a secondary mode located at 60 ppbv. The secondary mode is observed when RH ≤ 40%, agreeing with the results from Pan et al.(2015). Reanalysis data from MERRA-2 shows a primary and secondary mode of ozone concentrations at 20 ppbv and 40 ppbv, respectively. The secondary mode is found when RH ≤ 80%. These results indicate a relationship between high ozone concentration and low water vapor mixing ratio.

In addition to examining the relationships between ozone, cloud presence, and water vapor, we also investigated the relationships between ozone, water vapor, and vertical velocity. A vertical profile of ozone, colored by vertical velocity, shows positive vertical velocities (updrafts) between 300 hPa and 450 hPa, as well as below 700 hPa. Between 500 hPa and 700 hPa, and above 250 hPa, negative vertical velocities (downdrafts) are observed. This pattern was not found in the MERRA-2 reanalysis. In addition, enhanced ozone concentrations (above 100 ppbv) are observed when 0% < RH ≤ 20% and are associated with negative vertical velocities. The reanalysis data from MERRA-2 shows enhanced ozone concentrations when 0% < RH ≤ 20%, which are mainly associated with negative vertical velocities. Analysis was performed using vertical profiles of ozone-CO ratios using the CONTRAST campaign data. These profiles show a large difference between ozone-CO ratios when comparing different RH ranges (0% < RH ≤ 20%, 20% < RH ≤ 40%, 40% < RH ≤ 60%, 60% < RH ≤ 80%, 80% < RH ≤ 100%, and 100% < RH). A smaller difference is observed when comparing “in-cloud” and “clear-sky” conditions for the in-situ ozone-CO vertical profiles. These differences once again highlight the relationship between ozone and water vapor. Lastly, more analysis comparing the data from the CONTRAST campaign and the MERRA-2 reanalysis data will be done. This will help provide better ozone, water vapor, and vertical velocity representations in numerical models.