Thunderstorm Electrification: An overview of recent observational results

Paul Krehbiel Langmuir Laboratory, Geophysical Research Center New Mexico Institute of Mining and Technology Socorro, New Mexico 87801 USA

13th International Conference on Atmospheric Electricity Beijing, August 2007

Studying thunderstorm electrification processes Observational data: ● In-situ measurements (balloon soundings, aircraft penetrations) ● Remote sensing measurements (lightning mapping, dual-pol radar) Laboratory studies ● Continue to concentrate on non-inductive ice-ice charging (NIC) Computational simulations ● Increasingly powerful ● Highly useful means of testing laboratory results and proposed mechanisms against observations This talk: Focus on observational results obtained from remote sensing measurements of storm electrical structure and development

Lightning mapping (LMA) observations in a normallyelectrified convective storm (Thomas et al., 2001) Intracloud (IC) flash

Cloud-to-ground (CG) flash

+ _

_ +

Classic tripolar interior charge structure, showing the negative charge wellcorrelated with precipitation (graupel/small hail) at storm mid-levels, and a 4 km-deep neutral region in the upper dipole.

Correlation between lightning mapping observations and balloon E sounding measurements of storm charge structure Rust et al. 2005, Bird City storm, June 3-4, STEPS 2000

_ +

Similar correlation between lightning and E-sounding data also seen in New Mexico by Coleman et al., 2003.

How charge polarity determined for IC lightning flashes From asymmetries in the flash development (e.g. Shao and Krehbiel, 1996). ●

Normal IC: Initial VHF sources develop upward with time into positive charge (i.e., away from negative charge). ● Delayed onset of sources in negative charge region. ● K-changes at end of flash propagate from (-) to (+) charge regions ●

+ _

(Normal polarity intracloud flash)

How charge polarity determined for IC lightning flashes From asymmetries in the flash development (e.g. Shao and Krehbiel, 1996). ●

Normal IC: Initial VHF sources develop upward with time into positive charge (i.e., away from negative charge). ● Delayed onset of sources in negative charge region. ● K-changes at end of flash propagate from (-) to (+) charge regions ●

+ _

(Normal polarity intracloud flash)

How charge polarity determined for IC lightning flashes Inverted IC: Initial VHF sources develop down with time into positive charge (still away from negative charge). ● Delayed onset of sources in negative charge region. ● K-changes at end of flash propagate from (-) to (+) charge regions ●

_ +

(Normal polarity intracloud flash)

How charge polarity determined for IC lightning flashes Inverted IC: Initial VHF sources develop down with time into positive charge (still away from negative charge). ● Delayed onset of sources in negative charge region. ● K-changes at end of flash propagate from (-) to (+) charge regions ●

_ +

(Normal polarity intracloud flash)

Flashes Determining overall storm charge structure:

Z

Classify sequence of flashes (IC and CG) ●

time _+ +

Z

10-min time interval, ~150 flashes, 36,000 VHF sources ●

Plot density of sources of each polarity (+ orange, -- blue)

●

Normal-polarity (+,-,+) tripole structure, horizontally extensive charge regions. ●

Y

(Kyle Wiens Ph.D. Dissertation, 2005)

X

Z

Lightning-inferred charge structure and charging currents in a small storm over Langmuir Laboratory; July 31, 1999 I1 and I2 determined using a cylindrical charge model and estimated breakdown threshold values to match the IC and CG flashing rates. Charging current densities: J1 ~ 50 mA/km2 (upper dipole) J2 ~ 15 mA/km2 (lower dipole) ● Both appear to be explained by noninductive ice-ice charging (NIC). ●

Positive graupel charging apparently weaker than negative.

●

(Krehbiel et al., 2004) ●

Similar results obtained in Florida as in New Mexico and other places

Some storms can begin with lower dipole, and with --CG instead of IC (e.g., Shao and Liu, 1987; Qie et al., 2005; Bruning et al., 2007) ●

Lightning-inferred charge structure and charging currents in a small storm over Langmuir Laboratory; July 31, 1999 I1 and I2 determined using a cylindrical charge model and estimated breakdown threshold values to match the IC and CG flashing rates. Charging current densities: J1 ~ 50 mA/km2 (upper dipole) J2 ~ 15 mA/km2 (lower dipole) ● Both appear to be explained by noninductive ice-ice charging (NIC). ●

Positive graupel charging apparently weaker than negative.

●

(Krehbiel et al., 2004) ●

Similar results obtained in Florida as in New Mexico and other places

Some storms can begin with lower dipole, and with --CG instead of IC (e.g., Shao and Liu, 1987; Qie et al., 2005; Bruning et al., 2007) ●

●

Everything okay, except for storms that produce +CG lightning ...

Results of studying +CG storms during STEPS 2000 Storms have an anomalous, invertedpolarity electrical structure. ●

●

Characterized by inverted IC flashes.

Develop deep mid-level positive charge and upper level negative charge. ●

_ +

+CG’s originate in mid-level (+) charge, relatively late in storm (or not at all).

●

(+) charge well-correlated with graupel and small hail, as if the graupel carries the charge. ●

●

Complex evolutionary structures.

Variety of electrical structures, different parts of +CG storm on 25 June, 2000 (Weiss et al., 2007)

a: Normal-polarity dipole, b: 4-level, double inverted dipole structure, c, d: 2-level, inverted dipole. +CG’s produced later in Sections B, C.

Initial lightning activity, 29 June 2000 tornadic storm

Hamlin et al., 2002 ●

Storm still relatively small; low lightning rates.

●

Electrical structure: Lower dipole of normal-polarity tripole.

●

Lower (+) charge in graupel, small hail.

●

Fully consistent with (+) charging of graupel by reversed-polarity NIC.

Early lightning activity, 3 June 2000 Bird City storm Initial lightning (first 10 min): inverted dipole aloft (top). (+) charge in graupel. ●

2257

_ +

Next 10+ min: Normal tripolelike structure (bottom), but not normal. No negative CG’s; lower positive charge dominant. ●

Could be consistent with elevated NIC charging, with lower charging current, I2, dominant. ●

2257-2300

+_ + Hamlin, 2004

As storm developed further, upper (+) went away, and convective part of storm developed deep mid-level (+) charge with upper (-) charge (next slide). ●

Fully inverted dipole in convective part of storm [deep mid-level positive charge (red), shallow upper negative (green)]. ● Transient higher positive charge in downwind anvil. ●

2316

Wiens, 2005

2331

Positive charge carried by graupel/small hail ! Hamlin et al., 2002 (23:13:39 - 23:14:00)

Three successive IC discharges into hail echo

(00;20 - 00;28)

Descending IC discharges through deep positive charge

No CG!!

Never see this kind of lightning behavior in negative charge region of normal-polarity storms.

Initial lightning evolution, 29 June tornadic storm: Rapidly develops 3 additional charge layers in upper part of storm! ● Two inverted dipoles with separate discharges in lower and upper dipole pairs, capped by uppermost (+) charge layer. ●

First 50 min of lightning-inferred charge, 29 June tornadic storm: Initial low-rate lightning in lower inverted dipole (magenta), 20 – 25 min. ●

4- to 5-layer dual-dipole charge structure, with independent lightning in the two dipoles (yellow). ●

_ + _

_ +

+

Difficult to see how the multilayer structure could be result of NIC charging.

●

Rapid onset of multiple layers indicative of some sort of positive feedback (inductive) charging process.

●

Multi-layer structure persists until later becoming single inverted dipole with deep (+). ●

+CG’s start 1.0 to 1.5 hours into storm ! ●

Wiens, 2005

More descending discharges through vertically extensive positive charge in graupel and (large) hail shafts

Hamlin et al., 2002

Descending lightning not just in convective part of storm, most is downwind of the core.

Wiens, 2005

What mechanism produces the downwind deep positive charge?

What does all this mean?? Some tentative conclusions/suggestions: (+) NIC charging of graupel and small hail is alive and well. - Produces lower dipole of normal-polarity storms, and in anomalous storms that go on to become inverted. - Likely responsible for elevated mid-level (+) charge at beginning of inverted polarity storms where initial electrification begins aloft. - Undoubtedly responsible to some extent for deep positive charge of fully developed and inverted anomalous storms.

●

NIC charging is unlikely to explain multilayer structures that are common features of anomalous storms. Some other mechanism(s) at work here.

●

Other mechanisms almost certainly at work in producing the dominant, deep positive charge region of fully developed inverted storms. Graupel and hail appears still to be carrying the (+) charge, however.

●

The essentially continuous lightning activity in the anomalous storms must itself play a role in the electrification process. A huge source of ions.

●

Lightning-induced electrification would become self-sustaining – a characteristic feature even of smaller storms. (Possible mechanism: Preferential

●

corona from graupel/hail in vicinity of lightning channels; Krehbiel, 1984).

What about normally electrified storms?? If (+) NIC charging is so strong and dominant, even at mid- and upper-storm levels, or if other charging mechanisms are involved in producing anomalous storms: ●

- Raises questions about whether negative NIC charging is responsible for main negative charge of normally electrified storms. - Negative NIC charging of graupel may get electrification started, but then get taken over by something else. - (Still do not know that (-) charge is actually carried by ice-form precipitation. ) - Rather than seeing vertically descending lightning through negative charge regions, see extensive constant-altitude horizontal propagation instead (!) ●

Is the above trying to tell us something ??

●

(See Poster PS 3-13 )

Possible charging processes: ●

Non-inductive collisional charging of graupel/hail (NIC +, -).

●

Inductive charging (or other positive feedback mechanisms)

●

Ion capture near lightning channels

●

Ice splintering during freezing/riming

●

Preferential corona from ice hydrometeors

●

(Other, unknown processes ??)

More questions: Significance of neutral region between main (-) and upper (+) in normally electrified storms, versus lack of a neutral region in fully inverted anomalous storms ●

Dual Polarization radar signature of freezing/frozen liquid drops; 2 minutes before initial lightning

ZH

ZDR

ZV

ˆHV

MHV

Zunpolarized

First lightning discharge in storm, soon after disappearance of signature. Initiated in graupel in upper part of storm, above region of frozen liquid drops

ZH

ZDR

ZV

ˆHV

MHV

. (Earlier examples: Jameson et al., 1996; Krehbiel et al., 1996; Bringi et al., 1997; Carey and Rutledge, 2000; Bruning et al., 2007.)

(The End)