Cloud-resolving chemistry simulation of a Hector thunderstorm

Abstract

Cloud chemistry simulations were performed for a Hector thunderstorm observed on 16 November 2005 during the SCOUT-O3/ACTIVE campaigns based in Darwin, Australia, with the primary objective of estimating the average NO production per lightning flash in this unique storm type which occurred in a tropical island environment. The 3-D WRF-Aqueous Chemistry (WRF-AqChem) model is used for these calculations and contains the WRF nonhydrostatic cloud-resolving model with online gas- and aqueous-phase chemistry and a lightning-NO_x (LNO_x) production algorithm. The model was initialized by inducing convection with an idealized morning sounding and sensible heat source, and initial condition chemical profiles from merged aircraft observations in undisturbed air. Many features of the idealized model storm, such as storm size and peak radar reflectivity, were similar to the observed storm. Tracer species, such as CO, used to evaluate convective transport in the simulated storm found vertical motion from the boundary layer to the anvil region was well represented in the model, with a small overestimate of enhanced CO at anvil altitudes. The lightning detection network (LINET) provided lightning flash data for the model and a lightning placement scheme injected the resulting NO into the simulated cloud. A lightning NO production scenario of 500 moles flash⁻¹ for both CG and IC flashes yielded anvil NO_x mixing ratios that compared well with aircraft observations and were also similar to those deduced for several convective modeling analyses in the midlatitudes and subtropics. However, these NO production values were larger than most estimates for tropical thunderstorms and given several uncertainties, LNO_x production may have been as large as 600 moles flash⁻¹. Approximately 85% of the simulated LNO_x mass was located above 7 km in the later stages of the storm, which was greater than amounts found for subtropical and midlatitude convection. Modeled upper tropospheric NO₂ partial columns were also considerably greater than most satellite observations of tropical marine convective events, as tropical island convection, such as Hector, is more vigorous and more productive of LNO_x. Additional research is needed to investigate whether LNO_x production per flash increases in storms with greater wind shear, such as this Hector storm, which showed significant variation in wind direction with altitude

Evaluating a lightning parameterization based on cloud-top height for mesoscale numerical model simulations

“…11). Similar underestimation of the w max formulation has been noted for both tropical (Hector storm near Darwin, Australia) and US continental storms (Cummings et al, 2013). These results indicate that it is important to evaluate the flash rate parameterizations with observations.…”

The impact of overshooting deep convection on local transport and mixing in the tropical upper troposphere/lower stratosphere (UTLS)

“…For example, Cummings et al (2013) see an overestimation of modelled CO mixing ratios at anvil altitudes of a simulated Hector. Zhu et al (2013) concluded that the imperfect simulations of convection may lead to underestimation of the exchange of trace gases between troposphere and stratosphere.…”

Sensitivity of tropical tropospheric composition to lightning NOx production as determined by replay simulations with GEOS‐5

“…Therefore, the factor-of-four uncertainty currently associated with the global LNO x source strength stems primarily from the uncertainty in the global mean NO x production per flash. Several methods have been used to estimate this quantity: theoretical estimates [Price et al, 1997;Koshak et al, 2014], laboratory experiments ], cloud-resolving model simulations constrained by lightning flash observations and anvil NO x measurements [DeCaria et al, 2005;Ott et al, 2007Ott et al, , 2010Cummings et al, 2013], analyses of aircraft data [Huntrieser et al, 2008[Huntrieser et al, , 2009[Huntrieser et al, , 2011, and analyses of satellite observations [Beirle et al, 2004[Beirle et al, , 2006[Beirle et al, , 2010Bucsela et al, 2010]. NO x production estimates from these studies vary from 30 to 1100 mol flash À1 , though most estimates fall between 100 and 500 mol flash…”