The Deep Convective Clouds and Chemistry (DC3) experiment seeks to understand the kinematic and microphysical controls on the lightning behavior of deep moist convection. This study utilized multiple dual-polarization Doppler radars across northern Alabama to quantify microphysical and kinematic properties and processes that often serve as precursors to lightning, such as the graupel echo volume, graupel mass, and convective updraft volume. The focus here was on one multicellular complex that occurred on 21 May 2012 in northern Alabama during DC3. The graupel echo volume and the graupel mass in the charging region correlated well with the total lightning flash rate (FR), and even better than the updraft volumes and maximum updraft velocities. The flash length scales (LS) and flash areas were generally anticorrelated to the FR, while it was correlated to the nonprecipitation ice volume. More specifically, the presence of smaller flashes was associated with a stronger lower positive charge region caused by larger graupel volumes, stronger updraft volumes, and stronger maximum updraft velocities while larger flashes occurred during lower FRs and were associated with a weakened lower positive charge region in combination with a stronger upper positive charge region, weaker updraft velocities, a smaller graupel volume and mass, and an increase in nonprecipitation ice volume.

Deep convective transport of gaseous precursors to ozone (O3) and aerosols to the upper troposphere is affected by liquid phase and mixed‐phase scavenging, entrainment of free tropospheric air and aqueous chemistry. The contributions of these processes are examined using aircraft measurements obtained in storm inflow and outflow during the 2012 Deep Convective Clouds and Chemistry (DC3) experiment combined with high‐resolution (dx≤3 km) WRF‐Chem simulations of a severe storm, an air mass storm, and a mesoscale convective system (MCS). The simulation results for the MCS suggest that formaldehyde (CH2O) is not retained in ice when cloud water freezes, in agreement with previous studies of the severe storm. By analyzing WRF‐Chem trajectories, the effects of scavenging, entrainment, and aqueous chemistry on outflow mixing ratios of CH2O, methyl hydroperoxide (CH3OOH), and hydrogen peroxide (H2O2) are quantified. Liquid phase microphysical scavenging was the dominant process reducing CH2O and H2O2 outflow mixing ratios in all three storms. Aqueous chemistry did not significantly affect outflow mixing ratios of all three species. In the severe storm and MCS, the higher than expected reductions in CH3OOH mixing ratios in the storm cores were primarily due to entrainment of low‐background CH3OOH. In the air mass storm, lower CH3OOH and H2O2 scavenging efficiencies (SEs) than in the MCS were partly due to entrainment of higher background CH3OOH and H2O2. Overestimated rain and hail production in WRF‐Chem reduces the confidence in ice retention fraction values determined for the peroxides and CH2O.

Cloud electrification leads to the production of nitrogen oxides (NO x), which has an effect on ozone concentrations. Currently large uncertainties exist regarding the contribution of lightning to the global and local NO x budget, even on a per flash basis. Most lightning NO x (LNO x) models distribute the LNO x at reflectivities (Z) ≥ 20 dBZ in the horizontal, while vertically, a Gaussian distribution function with a peak at-15 °C is used for cloud-to-ground (CG) flashes and a bimodal distribution function with peaks at-15 °C and-45 °C is used for inter-and intracloud (IC) flashes. This research aims to improve our basic understanding of lightning location relative to radar Z as a function of storm and flash type. Using data from the North Alabama Lightning Mapping Array (NALMA) and the Multi-Radar Multi-Sensor data suite, the results from analyzing a multicell storm, mesoscale convective system and supercell storm showed that 29.7%, 15.9% and 6.9% of all flashes initiated in regions where Z < 20 dBZ, respectively. The bimodal lightning initiation distribution for IC flashes was also not observed for any of the three storms. In addition, it is shown that when incorporating the propagation of the flash, the percentage of NALMA lightning sources located in regions where Z < 20 dBZ increases. Finally, when comparing flash types, the results show that Hybrid flashes have consistently larger sizes than IC and CG flashes, while IC and Hybrid flashes tend to have more sources located at Z < 20 dBZ than CG flashes.

This study investigates the kinematic and microphysical control of lightning properties, particularly those that may govern the production of nitrogen oxides (NOX = NO + NO2) via lightning (LNOX), such as flash rate, type, and extent. The NASA Lightning Nitrogen Oxides Model (LNOM) is applied to lightning observations following multicell thunderstorms through their lifecycle in a Lagrangian sense over Northern Alabama on 21 May 2012 during the Deep Convective Clouds and Chemistry (DC3) experiment. LNOM provides estimates of flash rate, type, channel length distributions, channel segment altitude distributions (SADs), and LNOX production profiles. The LNOM‐derived lightning characteristics and LNOX production are compared to the evolution of radar‐inferred updraft and precipitation properties. Intercloud, intracloud (IC) flash SAD comprises a significant fraction of the total (IC + cloud‐to‐ground [CG]) SAD, while increased CG flash SAD at altitudes >6 km occurs after the simultaneous peaks in several thunderstorm properties (i.e., total [IC + CG] and IC flash rate, graupel volume/mass, convective updraft volume, and maximum updraft speed). At heights <6 km, the CG LNOX production dominates the column‐integrated total LNOX production. Unlike the SAD, total LNOX production consists of a more equal contribution from IC and CG flashes for heights >6 km. Graupel volume/mass, updraft volume, and maximum updraft speed are all well correlated to the total flash rate (correlation coefficient, ρ ≥ 0.8) but are less correlated to total flash extent (ρ ≥ 0.6) and total LNOX production (ρ ≥ 0.5). Although LNOM transforms lightning observations into LNOX production values, these values are estimates and are subject to further independent validation.

Deep convective transport of surface moisture and pollution from the planetary boundary layer to the upper troposphere and lower stratosphere affects the radiation budget and climate. This study analyzes the deep convective transport in three different convective regimes from the 2012 Deep Convective Clouds and Chemistry field campaign: 21 May Alabama air mass thunderstorms, 29 May Oklahoma supercell severe storm, and 11 June mesoscale convective system (MCS). Lightning data assimilation within the Weather Research and Forecasting (WRF) model coupled with chemistry (WRF‐Chem) is utilized to improve the simulations of storm location, vertical structure, and chemical fields. Analysis of vertical flux divergence shows that deep convective transport in the 29 May supercell case is the strongest per unit area, while transport of boundary layer insoluble trace gases is relatively weak in the MCS and air mass cases. The weak deep convective transport in the strong MCS is unexpected and is caused by the injection into low levels of midlevel clean air by a strong rear inflow jet. In each system, the magnitude of tracer vertical transport is more closely related to the vertical distribution of mass flux density than the vertical distribution of trace gas mixing ratio. Finally, the net vertical transport is strongest in high composite reflectivity regions and dominated by upward transport.

In an effort to improve our knowledge on the horizontal and vertical distribution of lightning initiation and propagation, ~500 multicells (producing a total of 72,619 flashes), 27 mesoscale convective systems (producing 214,417 flashes) and 23 supercells (producing 169,861 flashes) that occurred over northern Alabama and southern Tennessee were analyzed using data from the North Alabama Lightning Mapping Array and the Multi‐Radar Multi‐Sensor suite. From this analysis, two‐dimensional (2‐D) histograms of where flashes initiated and propagated relative to radar reflectivity and altitude were created for each storm type. The peak of the distributions occurred between 8.0 and 10.0 km (−24.0 to −38.5 °C) and between 30 and 35 dBZ for flashes that initiated within multicellular storms. For flashes that initiated within mesoscale convective systems, these peaks were 8.0–9.0 km (−27.1 to −34.6 °C) and 30–35 dBZ, respectively, and for supercells, they were 10.0–12.0 km (−42.6 to −58.1 °C) and 35–40 dBZ, respectively. The 2‐D histograms for the flash propagations were slightly different than for the flash initiations and showed that flashes propagated in lower reflectivities as compared to where they initiated. The 2‐D histograms were also compared to test cases; the root‐mean‐square errors and the Pearson product moment correlation coefficient (R) were calculated with several of the comparisons having R values >0.7 while the root‐mean‐square errors were always ≤0.017 (≤10%), irrespective of storm type. Finally, the mean flash sizes for the multicell, mesoscale convective system, and supercell flashes were 8.3, 9.9, and 7.4 km, respectively.

Two dimensional (2‐D) histogram distributions of lightning flashes relative to radar reflectivity and altitude were created using a total of 41,180 intercloud/intracloud (IC) flashes, 3,326 cloud‐to‐ground (CG) flashes, and 4,349 hybrid (HY) flashes that originated in multicells; 111,479 IC flashes, 8,588 CG flashes, and 11,699 HY flashes that originated in mesoscale convective systems; and 91,283 IC flashes, 3,023 CG flashes, and 7,872 HY flashes that originated in supercells that occurred over northern Alabama and southern Tennessee. It was shown that although CG flashes initiate and propagate at the same altitude irrespective of storm type, IC flashes could have differences of up to 10 °C, while for HY flashes these differences increased to up to 20 °C relative to storm type. Further, IC, CG, and HY flashes propagated in lower reflectivities than where they initiated, while CG flashes initiated and propagated within higher reflectivities than IC and HY flashes. HY flashes were also twice as large as IC flashes and ~40% larger than CG flashes, and flashes that originated in mesoscale convective systems had larger overall sizes as compared to multicells and supercells. When comparing the new 2‐D histogram distributions to the legacy distributions used for the calculation of lightning‐produced nitrogen oxides (LNOx), it was shown that the new distributions perform much better, with higher Pearson product moment correlation coefficient values and much lower root‐mean‐square errors. These new distributions are thus more appropriate to use when modeling LNOx and will lead to more accurate LNOx estimations than using the legacy distributions.

No abstract