*3.2. Spatial Distributions of Trace Gas Anomalies during the 2015–16 El Niño Event*

Figure 5 shows the spatial distribution of O3 at 82 hPa in July 2015 and CO at 100 hPa in October 2015 as well as respective climatologies. Significant differences are noticed in the trace gas distributions in 2015 over the tropical region. Compared to the climatology, minimum O3 (100 ppbv) is noticed over the SEA and Atlantic regions but with a relatively zonally uniform feature in 2015. High values of CO (200 ppbv) are observed, particularly over SEA in 2015. These observations clearly indicate the huge difference between climatology and 2015. Figure 6 illustrates the spatial distribution of anomalies of O3 at 82 hPa and CO at 100 hPa with respect to the climatology in July and October 2015. Distinct characteristics in the trace gas anomalies are perceived between these two months. In the O3, strong negative anomalies (~200 ppbv) are observed over the entire tropical region except some parts of the WP region in July 2015. No such changes are observed for O3 in October 2015 over the tropical region. However, the observed O3 decrease is much higher and recorded strong decrease in its concentrations (~200 ppbv) in the recent decade. The O3 and temperatures in the Tropical Tropopause Layer (TTL) are linked both dynamically and radiatively. It is reported that the O3 perturbations have a positive radiative feedback, with negative O3 anomalies locally cooling the TTL and positive O3 anomalies locally warming the TTL (Gilford et al., 2016). We also observed the negative cold point tropopause temperature (CPT) anomalies from COSMIC GPS RO data in July and August 2015 over most of the tropical equatorial region except over SEA and WP regions (Figures are not shown). The observed CPT anomalies are well correlated with the O3 anomalies in July and August. Large anomalies of CO (~150 ppbv) are observed in October 2015 and the prominent enhancement is observed over SEA region. The observed increase of CO mixing ratio is much higher compared to the previous El Niño events in the 21st century. This clearly indicates the transport of lower troposphere air into the tropopause level in October 2015.

**Figure 5.** Background climatology (averaged 2006–2014) of (**a**) ozone mixing ratio at 82 hPa in July and (**b**) carbon monoxide at 100 hPa in October. (**c**,**d**) same as above but for the year 2015.

**Figure 6.** Spatial distribution of ozone anomalies observed at 82 hPa in (**a**) July 2015 and (**c**) October 2015. (**b**,**d**) same as (**a**,**c**), respectively but for carbon monoxide anomalies observed at 100 hPa. Anomalies are calculated by removing the background climatology (January 2006–December 2014) of individual months.

The 2015 El Niño induced drought conditions, which further allowed active biomass burning and forest fires to spread rapidly in the SEA region in September and October [18,19]. These carbon emissions, which occurred in 2015 are the largest emissions since 1997 [18]. The effects of these fires are clearly seen in the enhancement of CO concentrations in the UTLS region. Due to these emissions, CO at the tropopause level is increased in 2015. The observed values of decreasing O3 (200 ppbv) and increasing CO (~150 ppbv) are high values recorded in 2015 and the results clearly indicate the unusual strong enhancement/decrease of trace gases (CO and O3) in the UTLS region. The carbon emissions in September and October over maritime SEA play an important role in the enhancement of CO in the UTLS region. Based on ground based and satellite measurements, it is well reported that the strong enhancement of carbon emissions over Indonesia happened in September–October 2015 [18–20]. It is also evident that the overall emissions from the tropical Asian biomass burning in 2015 were almost three times the 2001–2014 average [54]. In the present study, we also observed the strong increase in the CO concentrations in most of the troposphere even up to the 100 hPa over SEA and WP regions in October 2015 (Figure 7). MLS data is available from 215 hPa and has only 4 pressure levels (215, 146,100 and 68 hPa) in the troposphere to the UTLS region. Based on MLS data, it is difficult to see the vertical change in the CO. To avoid this we have utilized AIRS observed monthly mean CO data over SEA region from January to December 2015. Height-time cross section of CO over the SEA and WP regions observed from AIRS measured CO data from January to December 2015 is shown in Figure 7. It is clear that the vertical transport of the CO into the UTLS region is clearly seen in the Height-time cross section of the CO. The anomalies in these trace gases clearly demonstrated the unusual changes that are occurring in the UTLS region before the El Niño becomes strong and strengthened during the 2015–16 winter. It was also clear from the study that localized strong carbon emissions over SEA play a crucial role on the large enhancement of zonal mean CO. In the 'State of the Climate 2015' [54], it is clearly shown that the biomass burning in Indonesia region led to increasing of the CO, aerosols and tropospheric O3 in 2015. Our satellite measurement results are also well matched with their results. In the next section, we explore the changes in tropopause temperature and WV in the LS during this event.

**Figure 7.** Pressure time cross section of monthly mean carbon monoxide observed over Southeast Asia and Western Pacific region (averaged over 10◦N–10◦S/85◦E–140◦E) from January to December 2015. To obtain this, AIRS satellite measured carbon monoxide data is utilized.
