How waviness in the circulation changes surface ozone: a viewpoint using local finite-amplitude wave activity

Published in Atmospheric Chemistry and Physics, 2019

Recommended citation: Sun, Wenxiu, Peter Hess, Gang Chen and Simone Tilmes, 2019: How waviness in the circulation changes surface ozone: a viewpoint using local finite-amplitude wave activity, Atmospheric Chemistry and Physics, 19, 12917--12933, doi:10.5194/acp-19-12917-2019.

ABSTRACT: Local finite-amplitude wave activity (LWA) mea- sures the waviness of the local flow. In this work we relate the anticyclonic part of LWA, AWA (anticyclonic wave activ- ity), to surface ozone in summertime over the US on interan- nual to decadal timescales. Interannual covariance between AWA diagnosed from the European Centre for Medium- Range Weather Forecast Era-Interim reanalysis and ozone measured at EPA Clean Air Status and Trends Network (CASTNET) stations is analyzed using maximum covariance analysis (MCA). The first two modes in the MCA analysis explain 84% of the covariance between the AWA and MDA8 (maximum daily 8 h average ozone), explaining 29% and 14% of the MDA8 ozone variance, respectively. Over most of the US we find a significant relationship between ozone at most locations and AWA over the analysis domain (24– 53◦ N and 130–65◦ W) using a linear regression model. This relationship is diagnosed (i) using reanalysis meteorology and measured ozone from CASTNET, or (ii) using meteo- rology and ozone simulated by the Community Atmospheric Model version 4 with chemistry (CAM4-chem) within the Community Earth System Model (CESM1). Using the linear regression model we find that meteorological biases in AWA in CAM4-chem, as compared to the reanalysis meteorology, induce ozone changes between −4 and +8 ppb in CAM4- chem. Future changes (ca. 2100) in AWA are diagnosed in different climate change simulations in CAM4-chem, simu- lations which differ in their initial conditions and in one case differ in their reactive species emissions. All future simula- tions have enhanced AWA over the US, with the maximum enhancement in the southwest. As diagnosed using the linear regression model, the future change in AWA is predicted to cause a corresponding change in ozone ranging between −6 and 6 ppb. The location of this change depends on subtle features of the change in AWA. In a number of locations this change is consistent with the magnitude and the sign of the overall simulated future ozone change.

Download paper: journal website