Ozone impact from solar energetic particles cools the polar stratosphere

Abstract

Understanding atmospheric impacts of solar energetic particle precipitation (EPP) remains challenging, from quantification of the response in ozone, to implications on temperature. Both are necessary to understand links between EPP and regional climate variability. Here we use a chemistry-climate model to assess the importance of EPP on late winter/spring polar stratosphere. In transient simulations, the impact on NO_y, ozone, and temperature is underestimated when using EPP forcing from the current recommendation of the Coupled Model Intercomparison Project (CMIP6). The resulting temperature response is largely masked by overall dynamical variability. An idealised experiment with EPP forcing that reproduces observed levels of NO_y results in a significant reduction of ozone (up to 25%), cooling the stratosphere (up to 3 K) during late winter/spring. Our results unravel the inconsistency regarding the temperature response to EPP-driven springtime ozone decrease, and highlight the need for an improved EPP forcing in climate simulations.

Fig. 1. Representative annual stratospheric variability.

Monthly mean anomalies in (a.) residual vertical wind, ${\overline{w}}^{*}$ (0.5 mm s⁻¹ contour interval), (b.) temperature, T (3 K contour interval), (c.) nitrogen family, NO_y (NO_x, HNO₃, N₂O₅, ClONO₂, HNO₄) (0.8 ppbv contour interval), and (d.) ozone, O₃ (0.2 ppmv contour interval), averaged over latitudinal range 60–90°S. Pressure levels on the y-axis are 0.5–80 hPa, with approximate altitude in km given on the right-hand side. Contour color scales for each panel are given on the right hand side.

Fig. 2. Chemical and dynamical heating contributions in the stratosphere.

Schematic diagram of the potential interactions between chemical and dynamical heating contributions to temperature T, as a result of changes to ozone (O₃) and up/downwelling. Red arrows indicate an increasing effect, while blue arrows indicate a reducing effect. EPP is energetic particle precipitation, NO_y is nitrogen family (NO_x, HNO₃, N₂O₅, ClONO₂, HNO₄), QRS is short-wave radiative heating, QRL is long-wave radiative cooling, ${\overline{w}}^{*}$ is the residual vertical wind anomaly and χ_i is mixing ratio of the radiatively reactive gas. Pathways ab and cd highlight the dynamical effect on temperature both directly, and via feedback to radiative cooling (marked with blue and red dotted lines).

Fig. 3. Stratospheric response to interannual variations in energetic particle precipitation (EPP).

Nitrogen family, NO_y (NO_x, HNO₃, N₂O₅, ClONO₂, HNO₄), ozone (O₃), short-wave heating rate (QRS), temperature (T) and vertical residual wind (${\overline{w}}^{*}$) for High EPP-Low EPP years in the SH polar region (60–90°S). Hatching indicates areas not statistically significant at 95% level based on the t test (p > 0.05). NO_y differences of 0.001 ppmv are marked with bold black contour. Pressure levels on the y-axis are 0.5–80 hPa, with approximate altitude in km given on the right-hand side. Contour intervals from left to right are 0.001 ppmv, 0.1 ppmv, 0.05 K day⁻¹, 0.5 K, and 0.05 mm s⁻¹. Contour color scales for each panel are given on the top of the panel.

Fig. 4. Connection between chemistry and dynamics.

Correlation of monthly mean polar cap nitrogen family, NO_y (NO_x, HNO₃, N₂O₅, ClONO₂, HNO₄) and ozone (O₃), shortwave heating rate (QRS), temperature (T) anomalies for (a.). High energetic particle precipitation (EPP) years, and (b.) Low EPP years. Last panels shows the correlation between temperature and ${\overline{w}}^{*}$ anomalies. Hatching indicates areas not statistically significant at 95% level (p > 0.05). NO_y differences of 0.001 ppmv from Fig. 3 are marked with bold black contour. Ozone differences of −0.2 ppmv from Fig. 3 are marked with magenta contour. Pressure levels on the y-axis are 0.5–80 hPa, with approximate altitude in km given on the right-hand side. Contour interval is 0.2. Contour color scales (same for all panels) are given on the right-hand side.

Fig. 5. Time-slice sensitivity experiment.

Monthly mean differences of SH polar (60–90°S) nitrogen family, NO_y (NO_x, HNO₃, N₂O₅, ClONO₂, HNO₄), ozone (O₃), shortwave heating rate (QRS), temperature (T) and residual vertical wind (${\overline{w}}^{*}$) (from left to right) for the idealized experiments labeled FW^max and FW^min: (a.) FW^max-FW^min, (b.) FW^max-FW^min for years with strong negative residual vertical wind differences, ${\overline{w}}^{*}$ (downwelling), and (c.) FW^max-FW^min for years with strong positive residual vertical wind differences, ${\overline{w}}^{*}$ (upwelling). Hatching indicates areas not statistically significant at 95% level based on the t test. The number of years used for computing the composite differences: (a) 50/50 years, (b, c) 10/10 years. NO_y differences of 0.003 ppmv are marked with bold black contour. Pressure levels on the y-axis are 0.5–80 hPa, with approximate altitude in km given on the right-hand side. Contour intervals for the columns from left to right are 0.003 ppmv, 5%, 0.1 K day⁻¹, 0.5 K (a) and 1 K (b, c), and 0.2 mm s⁻¹. Contour color scales for each column of panels are given at the top.