skip to main content


The NSF Public Access Repository (NSF-PAR) system and access will be unavailable from 11:00 PM ET on Thursday, June 13 until 2:00 AM ET on Friday, June 14 due to maintenance. We apologize for the inconvenience.

Title: Radar Reflectivity and Altitude Distributions of Lightning Flashes as a Function of Three Main Storm Types

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 dBZfor 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 havingRvalues >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.

more » « less
Award ID(s):
Author(s) / Creator(s):
Publisher / Repository:
DOI PREFIX: 10.1029
Date Published:
Journal Name:
Journal of Geophysical Research: Atmospheres
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    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.

    more » « less
  2. Abstract

    Global satellite studies show a maximum in deep convection and lightning downstream of the Andes in subtropical South America. The Remote sensing of Electrification, Lightning, And Mesoscale/microscale Processes with Adaptive Ground Observations (RELAMPAGO) field campaign was designed to investigate the physical processes that contribute to the rapid development of deep convection and mesoscale convective systems (MCSs) in Argentina. A lightning mapping array (LMA) was deployed to Argentina as part of RELAMPAGO to collect lightning observations from extreme storms in the region. This study combines lightning data from the LMA and the Geostationary Lightning Mapper onboardGOES‐16with 1‐km gridded radar data to examine the electrical characteristics of a variety of convective storms throughout their life cycle observed during RELAMPAGO. Results from the full campaign show 48% of flashes are associated with deep convection that occurs along the eastern edge of the Sierras de Córdoba (SDC) overnight. These flashes are 65 km2smaller on average compared to stratiform flashes, which occur most frequently 50–100 km east of the SDC in the early morning hours, consistent with the upscale growth of MCSs off the terrain. Analysis of the 13–14 December MCS shows that sharp increases in flash rates correspond to deep and wide convective cores that have high graupel and hail mass, 35‐dBZ volume, and ice water path. This work validates previous satellite studies of lightning in the region, but also provides higher spatial and temporal resolution information across the convective life cycle that has not been available in previous studies.

    more » « less
  3. Abstract

    A supercell thunderstorm formed as part of a cluster of severe storms near Kingfisher, Oklahoma on May 29, 2012 during the Deep Convective Clouds and Chemistry field experiment. This storm produced 5 hail, an EF‐1 tornado, and copious lightning over the course of a few hours within range of the Oklahoma Lightning Mapping Array and the KTLX WSR‐88D radar. This study focuses on a ∼1‐h interval during which a line of secondary convection formed and intensified within the anvil of the Kingfisher supercell. An analysis of radar reflectivity, radial velocity, and low‐level divergence shows that the formation of the secondary convection was consistent with a previously proposed mechanism; the instability leading to the convection was initiated by diabatic cooling in dry air below the anvil from sublimation, melting and evaporation of virga falling from the anvil, coincident with weak rising motion above a surface outflow boundary adjacent to the sub‐anvil downdraft. Prior to the formation of the secondary convection, flashes extended up to 60 km from the deep convection into the anvil. After the line of secondary convection formed, it initiated long lightning flashes that propagated along the line, and it continued to produce lightning as it moved eastward out from under the anvil. The charge structure inferred from flashes in the parent storm and in the downstream anvil suggests that charge was separated locally in the anvil following the development of an outflow boundary by the Kingfisher storm, which also contributed to the initiation of the secondary convection.

    more » « less
  4. Abstract. Deployed on the mountainous island of Corsica for thunderstormmonitoring purposes in the Mediterranean Basin, SAETTA is a network of 12 LMA(Lightning Mapping Array, designed by New Mexico Tech, USA) stations thatallows the 3-D mapping of very high-frequency (VHF) radiation emitted by cloud discharges in the60–66 MHz band. It works at high temporal (∼40 ns in each 80 µs time window) and spatial (tens of meters at best) resolutionwithin a range of about 350 km. Originally deployed in May 2014, SAETTA wascommissioned during the summer and autumn seasons and has now been permanentlyoperational since April 2016 until at least the end of 2020. We firstevaluate the performances of SAETTA through the radial, azimuthal, andaltitude errors of VHF source localization with the theoretical model ofThomas et al. (2004). We also compute on a 240 km × 240 km domainthe minimum altitude at which a VHF source can be detected by at least sixstations by taking into account the masking effect of the relief. We thenreport the 3-year observations on the same domain in terms of number oflightning days per square kilometer (i.e., total number of days during whichlightning has been detected in a given 1 km square pixel) and in terms oflightning days integrated across the domain. The lightning activity is firstmaximum in June because of daytime convection driven by solar energy input,but concentrates on a specific hot spot in July just above the intersectionof the three main valleys. This hot spot is probably due to the low-levelconvergence of moist air fluxes from sea breezes channeled by the threevalleys. Lightning activity increases again in September due to numeroussmall thunderstorms above the sea and to some high-precipitation events.Finally we report lightning observations of unusual high-altitude dischargesassociated with the mesoscale convective system of 8 June 2015. Most ofthem are small discharges on top of an intense convective core duringconvective surges. They are considered in the flash classification of Thomaset al. (2003) to be small–isolated and short–isolated flashes. The other high-altitude discharges, much less numerous, are long-range flashes that developthrough the stratiform region and suddenly undergo upward propagationstowards an uppermost thin layer of charge. This latter observation isapparently consistent with the recent conceptual model of Dye and Bansemer (2019) that explains such an upper-level layer of charge in the stratiformregion by the development of a non-riming ice collisional charging in amesoscale updraft. 
    more » « less
  5. Hail-bearing storms produce substantial socioeconomic impacts each year, yet challenges remain in forecasting the type of hail threat supported by a given environment and in using radar to estimate hail sizes more accurately. One class of hail threat is storms producing large accumulations of small hail (SPLASH). This paper presents an analysis of the environments and polarimetric radar characteristics of such storms. Thirteen SPLASH events were selected to encompass a broad range of geographic regions and times of year. Rapid Refresh model output was used to characterize the mesoscale environments associated with each case. This analysis reveals that a range of environments can support SPLASH cases; however, some commonalities included large precipitable water (exceeding that day’s climatological 90th-percentile values), CAPE < 2500 J kg−1, weak storm-relative wind speeds (<10 m s−1) in the lowest few kilometers of the troposphere, and a weak component of the storm-relative flow orthogonal to the 0–6-km shear vector. Most of the storms were weak supercells that featured distinctive S-band radar signatures, including compact (<200 km2) regions of reflectivity factor > 60 dB Z, significant differential attenuation evident as negative differential reflectivity extending downrange of the hail core, and anomalously large specific differential phase KDP. The KDPvalues often approached or exceeded the operational color scale’s upper limit (10.7° km−1); reprocessing the level-II data revealed KDP>17° km−1, the highest documented in precipitation at S band. Electromagnetic scattering calculations using the T-matrix method confirm that large quantities of small melting hail mixed with heavy rain can plausibly explain the observed radar signatures.

    more » « less