Understanding recent tropospheric ozone trends in the context of large internal variability: a new perspective from chemistry-climate model ensembles

Abstract

Observational records of meteorological and chemical variables are imprinted by an unknown combination of anthropogenic activity, natural forcings, and internal variability. With a 15-member initial-condition ensemble generated from the CESM2-WACCM6 chemistry-climate model for 1950-2014, we extract signals of anthropogenic (‘forced’) change from the noise of internally arising climate variability on observed tropospheric ozone trends. Positive trends in free tropospheric ozone measured at long-term surface observatories, by commercial aircraft, and retrieved from satellite instruments generally fall within the ensemble range. CESM2-WACCM6 tropospheric ozone trends are also bracketed by those in a larger ensemble constructed from five additional chemistry-climate models. Comparison of the multi-model ensemble with observed tropospheric column ozone trends in the northern tropics implies an underestimate in regional precursor emission growth over recent decades. Positive tropospheric ozone trends clearly emerge from 1950 to 2014, exceeding 0.2 DU yr-1 at 20-40N in all CESM2-WACCM6 ensemble members. Tropospheric ozone observations are often only available for recent decades, and we show that even a two-decade record length is insufficient to eliminate the role of internal variability, which can produce regional tropospheric ozone trends oppositely signed from ensemble mean (forced) changes. By identifying regions and seasons with strong anthropogenic change signals relative to internal variability, initial-condition ensembles can guide future observing systems seeking to detect anthropogenic change. For example, analysis of the CESM2-WACCM6 ensemble reveals year-round upper tropospheric ozone increases from 1995-2014, largest at 30S-40N during boreal summer. Lower tropospheric ozone increases most strongly in the winter hemisphere, and internal variability leads to trends of opposite sign (ensemble overlaps zero) north of 40N during boreal summer. This decoupling of ozone trends in the upper and lower troposphere suggests a growing prominence for tropospheric ozone as a greenhouse gas despite regional efforts to abate warm season ground-level ozone.

Diverging Ozone Trends Above Western North America: Boundary Layer Decreases Versus Free Tropospheric Increases

“…For seasonal trends, the largest difference in trends is found in JJA, followed by MAM, and similar trends are only found in SON. This finding adds to the growing body of evidence that surface trends are frequently disconnected from the general increases observed in the free troposphere (Cooper et al., 2020), as assessed by IPCC (Gulev et al., 2021), and also reproduced by a recent chemistry‐climate model simulation (Fiore et al., 2022).…”

“…Results from a further sensitivity analysis are presented in Figure S12 in Supporting Information to investigate the percentile trend estimates and uncertainties based on different starting years (1995, 1998, 2000, and 2005) and ending years (2019 and 2021). The year 1998 is selected as it represents a well‐known enhanced ozone episode that was observed across the Northern Hemisphere, and has been recently attributed to extreme biomass burning (Fiore et al., 2022). In general, the big picture remains similar, except that, (a) the lower percentile trends above California are shifted from positive over 1995–2021 toward negative over 2005–2021 (as discussed above, lower percentile trends in California are heavily affected by the pandemic years); and (b) if the estimates are based on the starting year in 1998 with unusually high ozone, its impact can be observed mainly on the reduction of trends at the higher percentiles above WNA, but the sensitivity is weak at the lower percentiles.…”

“…Results from a further sensitivity analysis are presented in Figure S12 in Supporting Information S1 to investigate the percentile trend estimates and uncertainties based on different starting years (1995, 1998, 2000, and 2005) and ending years (2019 and 2021). The year 1998 is selected as it represents a well-known enhanced ozone episode that was observed across the Northern Hemisphere, and has been recently attributed to extreme biomass burning (Fiore et al, 2022).…”

Origin, importance, and predictive limits of internal climate variability

“…This suggests that a perspective focused on risk and robust decision making (Sutton 2019, Mankin et al 2020, Reed et al 2022 is more pertinent to users of climate model information than the focus on a 'best-estimate' . This is increasingly being recognized by application communities dependent on climate science, such as climate-related health impacts (Garcia-Menendez et al 2017, Fiore et al 2022 or water resource management (Harding et al 2012, Chegwidden et al 2019, Smith et al 2022. A current example is the scoping of new guidelines for the management of the Colorado River in the Western U.S.…”

Changing PM2.5 and related meteorology over India from 1950–2014: a new perspective from a chemistry-climate model ensemble

“…These ensemble members have different initial conditions for all Earth system model components (ocean, atmosphere, etc.). Additional ensemble members are generated from each of these three initial conditions by slightly perturbing the atmosphere initial conditions (more details about the ensemble can be found in Fiore et al [37]; our current ensemble is 12 members). These ensemble members represent 12 different possibilities for trends in atmospheric composition and meteorological variables from 1950 to 2014, with each ensemble member representing a climate system state that could have arisen under the same forcing scenario.…”