Statistical response of middle atmosphere composition to solar proton events in WACCM-D simulations: the importance of lower ionospheric chemistry

Abstract

Abstract. Atmospheric effects of solar proton events (SPEs) have been studied for decades, because their drastic impact can be used to test our understanding of upper stratospheric and mesospheric chemistry in the polar cap regions. For example, odd hydrogen and odd nitrogen are produced during SPEs, which leads to depletion of ozone in catalytic reactions, such that the effects are easily observed from satellites during the strongest events. Until recently, the complexity of the ion chemistry in the lower ionosphere (i.e., in the D region) has restricted global models to simplified parameterizations of chemical impacts induced by energetic particle precipitation (EPP). Because of this restriction, global models have been unable to correctly reproduce some important effects, such as the increase in mesospheric HNO3 or the changes in chlorine species. Here we use simulations from the WACCM-D model, a variant of the Whole Atmosphere Community Climate Model, to study the statistical response of the atmosphere to the 66 strongest SPEs which occurred in the years 1989–2012. Our model includes a set of D-region ion chemistry, designed for a detailed representation of the atmospheric effects of SPEs and EPP in general. We use superposed epoch analysis to study changes in O3, HOx (OH + HO2), Clx (Cl + ClO), HNO3, NOx (NO + NO2) and H2O. Compared to the standard WACCM which uses an ion chemistry parameterization, WACCM-D produces a larger response in O3 and NOx and a weaker response in HOx and introduces changes in HNO3 and Clx. These differences between WACCM and WACCM-D highlight the importance of including ion chemistry reactions in models used to study EPP.

Odd hydrogen response thresholds for indication of solar proton and electron impact in the mesosphere and stratosphere

“…The instrument is described in detail by Waters et al (2006). A detailed description of MLS version 4 data products and quality is given by Livesey et al (2018). The MLS target species in the stratosphere and mesosphere include OH and HO 2 .…”

“…In the analysis, we use the pressure level grid from the MLS measurements, as sections of the MLS HO 2 and OH pressure grids are the same. Hence, the overlapping sections of MLS HO 2 and OH values were selected, whereas the least squares interpolation method recommended by Livesey et al (2018) was used to convert WACCM-D data onto the MLS grid. The same pressure grid allows for direct comparisons between the data sets.…”

“…However, the noisiness of MLS HO 2 data is evident at lower atmospheric levels. Note that the HO 2 data could be improved below the 1 hPa level by using the daynight difference (Livesey et al, 2018), but this was not done in our analysis.…”

“…Compared to the OH observations, MLS HO 2 measurements have been made since 2004 and, thus, provide a longer time series than the OH observations. However, the standard HO 2 data from operational processing have a lower signal-to-noise ratio and only cover the lower mesosphere and stratosphere (Pickett et al, 2008;Livesey et al, 2018). Thus, the use of standard HO 2 data has been limited to large proton events (Jackman et al, 2011;Zou et al, 2018).…”

Response to the comments of Referee #2

“…We have highlighted this matter in our conclusions and also discuss the use of other species in the detection of EPPs in the revised manuscript. Kalakoski et al (2020) showed SPE-related increases in WACCM-D Cl x down to 1 hPa lasting around a week following an event. They also showed a response in HNO 3 above 1 hPa with slightly C5…”