A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1

Nicely, Julie M.; Duncan, Bryan N.; Hanisco, Thomas F.; Wolfe, Glenn M.; Salawitch, Ross J.; Deushi, Makoto; Haslerud, Amund S.; Jöckel, Patrick; Josse, Béatrice; Kinnison, Douglas E.; Klekociuk, Andrew; Manyin, Michael E.; Marécal, Virginie; Morgenstern, Olaf; Murray, Lee T.; Myhre, Gunnar; Oman, Luke D.; Pitari, Giovanni; Pozzer, Andrea; Quaglia, Ilaria; Revell, Laura E.; Rozanov, Eugene; Stenke, Andrea; Stone, Kane; Strahan, Susan; Tilmes, Simone; Tost, Holger; Westervelt, Daniel M.; Zeng, Guang

The hydroxyl radical (OH) plays critical roles within the troposphere, such as determining the lifetime of methane (inline-formulaCH4), yet is challenging to model due to its fast cycling and dependence on a multitude of sources and sinks. As a result, the reasons for variations in OH and the resulting methane lifetime (inline-formula M2inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimg05ab45943081ffb3661d7ec9bd6fac87 acp-20-1341-2020-ie00001.svg22pt12ptacp-20-1341-2020-ie00001.png ), both between models and in time, are difficult to diagnose. We apply a neural network (NN) approach to address this issue within a group of models that participated in the Chemistry-Climate Model Initiative (CCMI). Analysis of the historical specified dynamics simulations performed for CCMI indicates that the primary drivers of inline-formula M3inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimg6cff8140caa772a7f53e45699d7c8dd3 acp-20-1341-2020-ie00002.svg22pt12ptacp-20-1341-2020-ie00002.png differences among 10 models are the flux of UV light to the troposphere (indicated by the photolysis frequency inline-formulaJO1D), the mixing ratio of tropospheric ozone (inline-formulaO3), the abundance of nitrogen oxides (inline-formula M6inlinescrollmathml chem normal NO x normal NO + normal NO normal 2 85pt13ptsvg-formulamathimg219e22fa86429804f8520706e3902cdc acp-20-1341-2020-ie00003.svg85pt13ptacp-20-1341-2020-ie00003.png ), and details of the various chemical mechanisms that drive OH. Water vapour, carbon monoxide (CO), the ratio of inline-formulaNO:NOx, and formaldehyde (HCHO) explain moderate differences in inline-formula M8inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimg3ed8c39ea81cfa47f8a86470c8c7d2ad acp-20-1341-2020-ie00004.svg22pt12ptacp-20-1341-2020-ie00004.png , while isoprene, methane, the photolysis frequency of inline-formulaNO2 by visible light (inline-formulaJNO2), overhead ozone column, and temperature account for little to no model variation in inline-formula M11inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimge27a3d7e184809f515d99dd29b4c3358 acp-20-1341-2020-ie00005.svg22pt12ptacp-20-1341-2020-ie00005.png . We also apply the NNs to analysis of temporal trends in OH from 1980 to 2015. All models that participated in the specified dynamics historical simulation for CCMI demonstrate a decline in inline-formula M12inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimg5416400bfdb425bfb041111ba9583291 acp-20-1341-2020-ie00006.svg22pt12ptacp-20-1341-2020-ie00006.png during the analysed timeframe. The significant contributors to this trend, in order of importance, are tropospheric inline-formulaO3, inline-formulaJO1D, inline-formulaNOx, and inline-formulaH2O, with CO also causing substantial interannual variability in OH burden. Finally, the identified trends in inline-formula M17inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimgd40ce9ef1b678a5d0b2ee7f985bacba6 acp-20-1341-2020-ie00007.svg22pt12ptacp-20-1341-2020-ie00007.png are compared to calculated trends in the tropospheric mean OH concentration from previous work, based on analysis of observations. The comparison reveals a robust result for the effect of rising water vapour on OH and inline-formula M18inlinescrollmathml italic τ chem normal CH normal 4 22pt12ptsvg-formulamathimg3c638831574feeb93ae0b6f1a98f1be1 acp-20-1341-2020-ie00008.svg22pt12ptacp-20-1341-2020-ie00008.png , imparting an increasing and decreasing trend of about 0.5 % decadeinline-formula−1, respectively. The responses due to inline-formulaNOx, ozone column, and temperature are also in reasonably good agreement between the two studies.



Nicely, Julie M. / Duncan, Bryan N. / Hanisco, Thomas F. / et al: A machine learning examination of hydroxyl radical differences among model simulations for CCMI-1. 2020. Copernicus Publications.


Rechteinhaber: Julie M. Nicely et al.

Nutzung und Vervielfältigung: