On the roles of deep convective clouds in tropospheric chemistry

Abstract

The impact of deep convection on tropospheric chemistry has been studied using a cloud-resolving model including integrated dynamics, cloud and aerosol microphysics, gaseous and aqueous chemistry, radiation, and lightning production of NO. Initialized with observational data over the tropical Pacific Ocean, sensitivity runs using a three-dimensional version of the model for short integration times (4 hours) and a two-dimensional version for relatively long integration times (30 hours) have been carried out. The impact of deep convection on tropospheric chemistry is found to occur not only through vertical transport but also by changing UV fluxes (and thus the photochemical rates), scavenging soluble species, and producing NO molecules through lightning which then collectively change the main chemical pathways. The formation of large amounts of ice as opposed to liquid water particles is shown to greatly change the efficiencies of gaseous and aqueous reactions. The results suggest that upward transport during deep convection can form a transient layer in the lower stratosphere with relatively high H2O concentrations. This increase in H2O together with enhancement of UV fluxes by upward reflection from the cloud anvils then produces high OH concentrations just above the cloud top. The existence of a deep convective tower and its associated anvils changes the gas phase production of most chemical species by more than a factor of 2. We find that the production of NO through lightning is very important to chemistry in the convective region. In the model runs including lightning, the reduced UV fluxes inside and below the convective tower and anvils and during the nighttime combine with large conversions of NO to NO2 and then to N2O5 and HNO3 to produce reductions of both Nox and O3 despite the massive lightning production of NO. This is because as many as 28% of the NOx molecules produced by lightning have been terminated through reactions with O3 and HOx before contributing to O3 production. Total O3 in the model domain is reduced by up to 11 % relative to the preconvective state. The lowest O3 mole fractions in the upper troposphere calculated in the simulations are slightly less than 2 ppb, which is an 80% decrease from the initial O3 mole fraction. Vertical profiles of O3 from model runs including lightning agree quite well with observations in the upper troposphere, implying that lightning-related O3 loss in cloud anvil regions can help explain the occasionally observed low O3 layers in the tropical upper troposphere. The study also indicates that during convection, only 9% of the dissolved SO2 is converted to sulfate so that the aqueous processes only contribute 21% to the total sulfate production inside the model domain (∼1000 km wide). This result is caused by the much lower solubility of SO2 relative to H2O2; the conversion of water to ice phase particles that terminates the aqueous chemistry; and the limited coverages and lifetimes of the liquid phase portions of the clouds. This result also suggests that the aqueous sulfate production calculated in global models which neglect the ice phase may be overestimated. Copyright 2000 by the American Geophysical Union.