Charge source of cloud-to-ground lightning and charge structure of a typical thunderstorm in the Chinese Inland Plateau

GPS-synchronized measurements of electric (E) field changes induced by lightning flashes were recorded at six stations in the northeastern verge of the Tibetan Plateau. The height and magnitude of charge neutralized by 65 return strokes, including 16 negative cloud-to-ground (CG) flashes and 2 positive CG flashes, have been fitted with the nonlinear least-square method based on the E field changes of CG flashes observed in a typical thunderstorm with larger-than-usual lower positive charge center (LPCC). Results show that the height of the charge region neutralized by negative CG flashes ranges from 3 km to 5 km above the ground, corresponding to an ambient temperature between − 2 °C and − 15 °C. For the two positive CG flashes, the neutralized charge regions are located at a height of about 5.5 km and the ambient temperature is about − 18 °C, indicating the existence of upper positive charge in the thunderstorm.     

The characteristics of cloud-to-ground lightning activity in hailstorms over northern China

The characteristics of cloud-to-ground (CG) lightning for ten hailstorms in Shandong Province of China were analyzed statistically. It is found that the hailstorms in this study present dominant positive CG flashes during periods of falling hail. One specific hailstorm on 16 June 2006 was studied in detail using the data from a CG lightning location network, Doppler radar and cloud images. Comparison between the brightness temperature of cloud-top and CG flash locations indicated that most flashes occurred in the region with temperatures lower than − 40 °C, while dense positive CG flashes occurred in the range between − 40 °C and − 50 °C. Negative CG flashes occurred mostly in the relative weak radar echo region, and positive CG flashes were distributed in the strong echo region especially with a large gradient of echo intensity. CG flashes tended to occur in the cloud region with reflectivity between 25 dBZ and 35 dBZ. Comparison between the wind field retrieved from Doppler radar and the location of CG flashes indicated that the flashes were located in the convergent region at lower to middle levels.     

On the generation of charge layers in MCS stratiform regions

We suggest a quantitative one-dimensional model treating the formation of charge layers near the 0 °C isotherm in stratiform regions of mesoscale convective systems. A number of factors principal for the field generation have been taken into account: both non-inductive and inductive melting charging, light ions, a complicated profile of the vertical air velocity near the 0 °С isotherm, the boundary conditions proper for the horizontally extended systems in the global electric circuit. Non-inductive collisional charging near the 0 °C isotherm was not considered. It was found that both non-inductive and inductive melting mechanisms can contribute; the inductive melting charging of ice aggregates was found more preferable, while the contribution of non-inductive mechanisms might be significant depending on particular conditions. The role of light ions in the formation of the positive charge layer near the 0 °C isotherm may be important. If the advection from the convective region ensures charge inflow to the upper charged layers, the melting charging mechanisms are able to explain an observable electric field structure in the whole stratiform region. It is important that the mutual position of the zero point on the vertical air velocity profile and the point of maximum melting-charge-transfer determines the fine structure of the electric field in the vicinity of the 0 °C isotherm.     

Cloud contribution to the daily and annual radiation budget in a mountainous valley

An automated-ventilated radiation station has been set up in a mountainous valley at the Logan Airport in northern Utah, USA, since mid-1995, to evaluate the daily and annual radiation budget components, and develop an algorithm to study cloudiness and its contribution to the daily and annual radiation. This radiation station (composed of pyranometers, pyrgeometers and a net radiometer) provides continuous measurements of downward and upward shortwave, longwave and net radiation throughout the year. The surface temperature and pressure, the 2-m air temperature and humidity, precipitation, and wind at this station were also measured. A heated rain gauge provided precipitation information. Using air temperature and moisture and measured downward longwave (atmospheric) radiation, appropriate formula (among four approaches) was chosen for computation of cloudless-skies atmospheric emissivity. Considering the additional longwave radiation during the cloudy skies coming from the cloud in the waveband which the gaseous emission lacks (from 8–13 μm), an algorithm was developed which provides continuous 20-min cloud information (cloud base height, cloud base temperature, percent of skies covered by cloud, and cloud contribution to the radiation budget) over the area during day and night. On the partly-cloudy day of 3 February, 2003, for instance, cloud contributed 1.34 MJ m^− 2 d^− 1 out of 26.92 MJ m^− 2 d^− 1 to the daily atmospheric radiation. On the overcast day of 18 December, 2003, this contribution was 5.77 MJ m^− 2 d^− 1 out of 29.38 MJ m^− 2 d^− 1. The same contribution for the year 2003 amounted to 402.85 MJ m^− 2 y^− 1 out of 9976.08 MJ m^− 2 y^− 1. Observations (fog which yielded a zero cloud base height and satellite cloud imaging data) throughout the year confirmed the validity of the computed data. The nearby Bowen ratio station provided the downward radiation and net radiation data. If necessary, these data could be substituted for the missing data at the radiation station. While the automated surface observing systems (ASOS) ceilometer at the Logan airport provides only the overhead cloud information, the proposed algorithm provides this information over the valley. The proposed algorithm is a promising approach for evaluation of the cloud base temperature, cloud base height, percent of skies covered by cloud, and cloud contribution to the daily and annual radiation budget at local and regional scales.     

Supercooled drizzle formed by condensation–coalescence in the mid-winter season of the Canadian Arctic

Supercooled drizzle (freezing drizzle) was observed at Inuvik, N.W.T., Canada (68°22′N, 133°42′W) on December 20, 21 and 27, 1995. Meteorological conditions in which the supercooled drizzle could form under low temperatures (colder than −20°C) in the mid-winter season of the Canadian Arctic were examined from the sounding data and data measured by a passive microwave radiometer at ground level. The following results were obtained. (1) Supercooled drizzle fell to the ground with ice pellets and frozen drops on snow crystals. (2) The maximum size of supercooled drizzle particles increased as the depth of cloud layer saturated with respect to water increased. (3) Because a layer of air temperature higher than 0°C was not detected from the sounding data at Inuvik, melting of snow particles was impossible. It was concluded, therefore, that supercooled drizzle was formed by the condensation–coalescence process below freezing temperature.     

Changes of cloud water chemical composition in the Western Sudety Mountains, Poland

The changing chemical composition of cloud water and precipitation in the Western Sudety Mountains are discussed against the background of air-pollution changes in the Black Triangle since the 1980s until September 2004. A marked reduction of sulphur dioxide emissions between the early 1990's and the present (from almost 2 million tons to around 0.2 million tons) has been observed, with a substantial decline of sulphate and hydrogen concentration in cloud water (SO₄²⁻ from more than 200 to around 70 μmol l^− 1; H⁺ from 150 to 50 μmol l^− 1) and precipitation (SO₄²⁻ from around 80 to 20–30 μmol l^− 1; H⁺ from around 60 to 10–15 μmol l^− 1) samples. At some sites, where fog/cloud becomes the major source of pollutants, deposition hot spots are still observed where, for example, nitrogen deposition can exceed 20 times the relevant critical load. The results show that monitoring of cloud water chemistry can be a sensitive indicator of pollutant emissions.     

The ice nucleating ability of pollen: Part III: New laboratory studies in immersion and contact freezing modes including more pollen types

Based on earlier experimental studies, the ice nucleating abilities of further pollen types were investigated in the immersion and contact freezing modes. The studies were carried out at the Mainz vertical wind tunnel with freely floating supercooled droplets down to − 28 °C. The pollen had diameters between 26 and 28 μm and correspondingly low sink velocities around 2.5 cm s^{− 1}. The radii of the studied drops were calculated from the recorded wind velocity and for both freezing modes the radii of the observed droplets varied between 315 and 380 μm. Immersion freezing experiments were conducted with pollen particles added to the droplets while in contact freezing experiments supercooled droplets were subjected to a burst of pollen particles. The median freezing temperatures found in the immersion freezing mode were: − 13.5 °C (alder), − 21.5 °C (lombardy poplar), − 21.0 °C (redtop grass) and − 19.8 °C (kentucky blue). The median freezing temperatures in the contact freezing mode were found as: − 12.6 °C (alder), − 17.9 °C (lombardy poplar), − 18.7 °C (redtop grass) and − 16.1 °C (kentucky blue). The results show that the ice nucleating ability of pollen is not restricted to single pollen types but seems to be a general pollen property.     

Effects of prescribed initial cloud droplet spectra on convective cloud and precipitation developments under different thermodynamic conditions: A modeling and observational study

A two-dimensional cloud model with bin microphysics was used to investigate the effects of cloud condensation nuclei (CCN) concentrations and thermodynamic conditions on convective cloud and precipitation developments. Two different initial cloud droplet spectra were prescribed based on the total CCN concentrations of maritime (300 cm^− 3) and continental (1000 cm^− 3) air masses, and the model was run on eight thermodynamic conditions obtained from observational soundings. Six-hourly sounding data and 1-hourly precipitation data from two nearby weather stations in Korea were analyzed for the year 2002 to provide some observational support for the model results.For one small Convective Available Potential Energy (CAPE) ( 300 J kg^− 1) sounding, the maritime and continental differences were incomparably large. The crucial difference was the production of ice phase hydrometeors in the maritime cloud and only water drops in the continental cloud. Ice phase hydrometeors and intrinsically large cloud drops of the maritime cloud eventually lead to significant precipitation. Meanwhile negligible precipitation developed from the continental cloud. For the three other small CAPE soundings, generally weak convective clouds developed but the maritime and continental clouds were of the same phases (both warm or both cold) and their differences were relatively small.Model runs with the four large CAPE ( 3000 J kg^− 1) soundings demonstrated that the depth between the freezing level (FL) and the lifting condensation level (LCL) was crucial to determine whether a cloud becomes a cold cloud or not, which in turn was found to be a crucial factor to enhance cloud invigoration with the additional supply of freezing latent heat. For two large CAPE soundings, FL–LCL was so deep that penetration of FL was prohibitive, and precipitation was only mild in the maritime clouds and negligible in the continental clouds. Two other soundings of similarly large CAPE had small FL–LCL, and both the maritime and continental clouds became cold clouds. Precipitation was strong for both but much more so in the maritime clouds, while the maximum updraft velocity and the cloud top were slightly higher in continental clouds. Although limited to small CAPE cases, more precipitation for smaller FL–LCL for a selected group of precipitation and thermodynamic sounding data from Korea was in support of these model results in its tendency.These results clearly demonstrated that the CCN effects on cloud and precipitation developments critically depended on the given thermodynamic conditions and not just the CAPE but the entire structure of the thermodynamic profiles had to be taken into account.     

Statistical analysis of microphysical properties and the parameterization of effective radius of warm clouds in Beijing area

In this paper warm cloud microphysical parameters including cloud droplet number concentration (N_c), liquid water content (q_l) and effective radius (r_e) from 75 flights around the Beijing area during 2005 and 2006 are summarized. Average N_c (cm^− 3) for Cu, Sc, Ac, As and Ns are 376 ± 290, 257 ± 226, 147 ± 112, 60 ± 35 and 60 ± 84, respectively. Many records of high N_c above 1000 cm^− 3 are observed. The large standard deviations indicate a large variation of N_c and q_l in this region. The maxima of q_l reach 1.4 g m^− 3 in Cu and 1.0 g m^− 3 in Sc, respectively. Different parameterizations of effective radius are examined with the in-situ data in this area. There are different ways to obtain the prefactor representing the relationship between effective radius and mean volume radius. Significant systematic errors are found to be at the large sizes when the prefactor is expressed with relative dispersion under the Gamma Distribution. Fixed prefactor of 1, which was widely used, even produces much larger error. A prefactor of 1.22 is found to be better than the former two methods by fitting with the observed data. The effective radius is further parameterized as functions of mean volume radius, liquid water content and cloud droplet number concentration. We suggest that the effective radius can be parameterized as r_e,p ≈ 1.20r_v + 0.22–1.28/r_v², which is a practical and more accurate scheme without too much computation complexity.     

One year measurements of aerosol optical properties over an urban coastal site: Effect on local direct radiative forcing

We present results of direct aerosol radiative forcing over a French Mediterranean coastal zone based on one year of continuous observations of aerosol optical properties during 2005–2006. Monthly-mean aerosol optical depth at 440 nm ranged between 0.1 and 0.34, with high Angstrom coefficient (α > 1.2). The single scattering albedo (at 525 nm) estimated at the surface ranged between 0.7 and 0.8, indicating significant absorption. The presence of aerosols over the Mediterranean zone during summer decreases the shortwave radiation reaching the surface by as much as 26 ± 3.9 W m^− 2, and increases the top of the atmosphere reflected radiation by as much as 5.2 ± 1.0 W m^− 2. The shortwave atmospheric absorption translates to an atmospheric heating of 2.5 to 4.6 K day^− 1. Concerted efforts are needed for investigating the possible impact of the increase in heating rate on the maintenance of heat-waves frequently occurring over this coastal region during summer time.     

Transient currents in the global electric circuit due to cloud-to-ground and intracloud lightning

Intracloud (IC) and cloud-to-ground (CG) lightning flashes produce transient changes in the electric field (E) above a thundercloud which drive transient currents in the global electric circuit (GEC). Using in-cloud and above-cloud E data from balloons, ground-based E data, and Lightning Mapping Array data, the above-cloud charge transfers due to lightning transients are estimated for five IC and five CG flashes from four thunderstorms that occurred above the mountains in New Mexico, USA, in 1999. For the five CG flashes (which transferred − 4 to − 13 C to the ground), the transient currents moved + 1 to + 5 C of charge upward from cloudtop toward the ionosphere, with an average transient charge transfer of about 35% of the charge transferred to ground. For the five IC flashes (which neutralized 6 to 21 C inside the cloud), the transient currents moved − 0.7 to − 3 C upward, with an average transient charge transfer of about 12% of the lightning charge. Estimates for three thunderstorms indicate that the transient currents made only a small GEC contribution compared to the quasi-stationary Wilson currents because of the offsetting effects of IC and CG flashes in these storms. However, storms with extreme characteristics, such as high flash rates or predominance of one flash type, may make a significant GEC contribution via lightning transients.     

Deposition of atmospheric particulate PCBs in suburban site of Turkey

Dry deposition and air concentration samples were collected from July 2004 to May 2005 at a suburban site in Turkey. A water surface sampler (WSS) was used to measure directly the dry deposition flux of particulate polychlorinated biphenyls (PCBs) while a high volume air sampler (HVAS) was employed to collect air samples. Particulate PCB concentrations accounted for 15% of total PCBs (gas + particle phase) at the site. The overall particulate phase PCB flux ranged from 2 to 160 ng m^− 2 d^− 1 with an average of 46.3 ± 40.6 ng m^− 2 d^− 1. Forty one PCB congeners were targeted in the samples while twenty one congeners were found to be higher than detection limits in deposition samples. Fluxes for homolog groups ranged between 0.9 (7-CBs) and 21.0 (3-CBs) ng m^− 2 d^− 1. Measured dry deposition fluxes were lower than the ones usually reported for urban sites. Average PCB dry deposition velocity, calculated using flux values and concurrently measured atmospheric concentrations, was 1.26 ± 1.86 cm s^− 1 depended on size distribution of particles, atmospheric PCB concentrations and meteorological conditions.     

Optical observations of transient luminous events associated with winter thunderstorms near the coast of Israel

We report the results of two observation campaigns conducted during the Northern Hemisphere winters of 2005–6 and 2006–7 aiming to detect transient luminous events (TLEs) above winter thunderstorms in the vicinity of Israel and the eastern coastline of the Mediterranean Sea. In 10 out of 31 different observation nights we detected 66 events: 56 sprites and 10 Elves. The detection ranges varied from 250 to 450 km. Sprites were found to be produced by active cells with a vertical dimension of 5–9 km and cloud top temperature ~ − 40 °C, embedded in a much larger matrix of stratiform precipitating cloudiness. This configuration closely resembles the conditions for winter sprites in the Hokuriku region of Japan. Synchronized with the optical observations, ELF data were recorded at two observation stations in Israel and Hungary in order to qualify and quantify parameters of the parent lightning discharge associated with the TLEs. These stations are located 500 km and 2100 km respectively from the Eastern Mediterranean Sea, where most TLEs occur. Among the optically observed TLE events, we found that all the ELF signals were produced solely by positive cloud-to-ground flashes (+ CGs), most of which were recorded in Israel (88%) and Hungary (77%). Calculation of the Charge Moment Change showed average values of 1400 ± 600 C km, with some extreme events exceeding 3500 C km. The average time delay between the ELF transient of the parent + CG and the observed sprites was 55 ms, with shorter delays for column sprites (42 ± 34 ms) compared to carrot sprites (68 ± 34). Furthermore, based on the ELF data, there were no early identifiable precursors to TLE occurrence in the regional lightning activity. From the spatial formation of the observed columniform sprites, we propose that columniform sprites are sometimes arranged in a 3-dimensional circular pattern, thus mapping the instantaneous electric field in the mesosphere.     

Cloud condensation nuclei measurements at Mace Head, Ireland, over the period 1994–2002

Analyses of cloud condensation nuclei (CCN) number concentrations (cm^− 3) measured at the Mace Head Atmospheric Research Station, near Carna, County Galway, Ireland, using a DH Associates Model M1 static thermal diffusion cloud chamber over the period from March 1994 to September 2002 are presented in this work. Air masses are defined as being ‘marine’ if they originate from a wind direction of 180–300° and ‘continental’ air masses are defined as originating from a wind direction of 45–135°. Air masses without such filtering were classified as ‘undefined’ air masses. Air masses were found to be dominated by marine sector air, re-affirming Mace Head as a baseline atmospheric research station. CCN levels for specific air masses at Mace Head were found to be comparable with earlier studies both at Mace Head and elsewhere. Monthly averaged clean marine (wind direction of 180–300° and black carbon absorption coefficient < 1.425 Mm^− 1) CCN and marine CCN varied between 15–247 cm^− 3 and 54–670 cm^− 3, respectively. As expected, significant increases in number concentration were found in continentally sourced CCN over that of marine CCN and were found to follow a log-normal distribution significantly tighter than that of clean marine air masses. No significant trend was found for CCN over the 9-year period. While polluted continental air masses showed a slight increase in CCN concentrations over the winter months, most likely due to increased fuel usage and a lower mixed boundary layer, the dominance of marine sector air arriving at Mace Head, which generally consists of background CCN concentrations, reduced seasonal differences for polluted air. Marine air showed a distinct seasonal pattern, with elevated values occurring over the spring and summer seasons. This is thought to be due to enhanced biogenic aerosol production as a result of phytoplankton bloom activity in the North Atlantic.     

The spatial and temporal variability of fog and its relation to fog oases in the Atacama Desert, Chile

Fog has been studied in the Atacama Desert of Chile for the past ten years. This paper analyzes its temporal and spatial variability, relying in part on satellite images (GOES) to analyze the frequent orographic fog and the low cloud deck (stratocumulus, Sc) that generates advective fog in the area. Fog water fluxes were measured with Standard Fog Collectors (SFC). Field trips and observers provided information on cloud top and base and the presence of fog. Vegetation in fog oases were used to confirm the results of these surveys.The Sc moves onshore into the continent with different intensities depending on season and time of day. The maximum spatial extent occurs during winter and at night. Fog is frequent in the coastal cliffs, where fog water fluxes of 7.0 L m^− 2 day^− 1 were measured using a SFC. It is less frequent 12 km inland, where the collection rates were less than 1 L m^− 2 day^− 1. The height of the fog collector above the ground affected the collection rate. The highest fog water fluxes were recorded at Alto Patache at altitudes of between 750 and 850 m a.s.l. The growth or thickness of the cloud is important in the collection of fog water. The information that GOES provides on the altitude of the top of low clouds is used to analyze this factor. Fog oases are described and analyzed in relation to how the geographical location of fog influences the growth of vegetation.     

The climate of the coast and fog zone in the Tarapacá Region, Atacama Desert, Chile

In the Atacama Desert, the narrow littoral plain and the adjacent mountain range have a unique climate. This area is locally called the “coastal desert with abundant cloudiness”, and extends from the coastline up to an elevation of 1000 m. The climate is designated as being BWn according to Köppen's Climate Classification as adapted for Chile. In the original classification the acronym (Bn) is used for foggy environments. Toward the east a “normal desert” climate (BW) is found. This is known as one of the most extreme deserts of the world. In the BWn areas there are meteorological differences between low and high elevation zones. The climate of the coastal plains and the mountains is described in this paper in order to show that there is an area where the climate differs from those classified as BWn and BW in the Chilean Climate Classification. This area is located between 650 and 1200 m a.s.l. and contains several fog oases or lomas vegetation, rich in biodiversity and endemism.The weather is warmer near sea level, with an annual average temperature of 18 °C. At high elevation sites like Alto Patache, the temperature decreases at a rate of 0.7 °C for every 100-m increase in altitude. The average annual minimum temperature often approaches 1 °C in winter, while the mean annual temperature range is significant (8.3 °C in Los Cóndores). The mean monthly relative humidity in Alto Patache is over 80%, except during the summer months. During autumn, winter and spring high elevation fog is present in the study area at altitudes ranging from 650 m up to 1060 m, giving annual water yields of 0.8 to 7 L m^− 2 day^− 1. If vegetation is used as an indicator, the foggy zone lies between 650 m a.s.l. and 1200 m a.s.l. About 70% of the mountain range experiences the foggy climate, as opposed to the coastal plains that are characterized by a cloudy climate.     

Ship measurements of submicron aerosol size distributions over the Yellow Sea and the East China Sea

During the spring of 2005, the total particle concentrations and the submicron aerosol size distributions were measured on board the research vessel over the south sea of Korea and the Korean sector of the Yellow Sea. Similar measurements were made over the East China Sea in autumn 2005. The aerosol properties varied dynamically according to the meteorological conditions, the proximity to the land masses and the air mass back trajectories. The average total particle concentration was the lowest over the East China Sea, 4335 ± 2736 cm^− 3, but the instantaneous minimum, 837 cm^− 3, for the entire ship measurement was recorded during the Yellow Sea cruise. There was also a long (more than 6 h) stretch of low total particle concentrations that fell as low as 1025 cm^− 3 during the East China Sea cruise when the ship was the farthest from the shores and the air mass back trajectories resided long hours over the sea. These observations lead to the suggestion of ~ 1000 cm^− 3 as the background total particle concentration over the marine boundary layer in the studied region of the Yellow Sea and the East China Sea, implying significant anthropogenic influence even for the background value. In the mean time, average aerosol size distributions were unimodal and the mode diameter ranged between 52 and 86 nm, excluding the fog periods, which suggests that the aerosols measured in this study experienced relatively less aging processes within the marine boundary layer.     

Study of the total lightning activity in a hailstorm

A thunderstorm that developed over northeastern Spain on 16 June 2006 is analyzed. This severe thunderstorm produced hailstones as large as 40 mm and had a lifetime of 3 h and 30 min. Radar cross-sections show strong vertical development with cloud echo tops reaching an altitude of 13 km. The specific characteristics of the lightning activity of this storm were: (i) a large amount (81%) of negative cloud-to-ground (−CG) flashes with very low peak currents (< 10 kA in absolute value), (ii) a very large proportion of intra-cloud (IC) flashes with an IC/CG ratio reaching about 400, (iii) a large number of “short” IC flashes (with only 1-VHF source according to SAFIR detection), (iv) a large increase of the −CG flash rate and of the CG proportion near the end of the storm. The rate of −CG flashes with a low peak current were observed to evolve similarly to the rates of IC flashes. Most of them have been assumed to be IC flashes misclassified by the Spanish Lightning Detection Network (SLDN). They have been filtered as it is usually done for misclassified +CG flashes. After this filtering, CG flash rates remained very low (< 1 min^− 1) with +CG flashes sometimes dominant. All the particular lightning activity characteristics similar to those observed in the Severe Thunderstorm Electrification and Precipitation Study (STEPS) campaigns support the hypothesis that this thunderstorm could have had an inverted-polarity or complex charge structure. The maximum IC flash rate (67 min^− 1) peaked 24 min before the presence of reflectivity higher than 60 dBZ. The IC activity abruptly decreased during the period when reflectivity was dramatically increasing. The time of maximum reflectivity observed by radar was consistent with the times of reported hail at the ground.     

Simultaneous optical and electrical observations on the initial processes of altitude-triggered negative lightning

Using a high-speed camera system and two electric field antenna systems, we have documented the initial processes of an altitude-triggered negative lightning (ATNL). The optical records clearly show that ATNL begins with the inception and propagation of an upward positive leader (UPL) and then a simultaneous propagation of UPL and downward negative leader (DNL), known as the bidirectional leader process, follows. Based on the optical records, it is inferred that (1) the triggering height is about 371 m; (2) the two-dimensional (2D) propagation speed of the UPL in its inception phase is about 3.8–5.5 × 10⁴ m s^− 1 during its propagation from about 393 to 452 m above the ground; (3) the grey levels of the DNL are about one order of magnitude higher than that of the UPL in their inception phase; (4) a discharge phenomenon propagating along the elevated triggering wire part of the lightning channel occurs after the mini-return stroke (MRS), with a 2D propagation speed of about 1.6–2.0 × 10⁵ m s^− 1. Combined with the simultaneous electric field change records, it is further inferred that (1) the UPL incepts about 932 μs earlier than the unstable DNL and about 4.1 ms earlier than the stable DNL; (2) the unstable DNL propagates downward intermittently three times with a time interval of about 1 ms, and each propagation contains a different number of steps with an average step length of about 7 m; (3) the stable DNL incepts at the tip of the unstable one, with a 2D propagation speed of about 1.9 × 10⁵ m s^− 1, an average step length of about 3 m, and a stepping time interval varying from 6 to 31 µs with a mean value of 15 µs.