Thunderstorm Charge Separation - PowerPoint Presentation

Slide1

Thunderstorm Charge Separation

Authors: S. E. Reynolds, M. Brook, and Mary

Foulks

Gourley

Published: October 1957

Presented By: Julie BarnumSlide2

Introduction – Development of Initial T-storm

Knew that lightning was formed only in convection reaching well above 0°C and could only happen in clouds that are precipitating

Looking at the earlier stages of a thunderstorm, as the system is earlier to understand as a whole

Development of initial cells in NM (Workman and Reynolds, 1949)

Initial radar echo from precipitation generally occurs at ~ -10°C (23,000 ft.), about 8,000 ft. below cloud top, near the vertical axis of rising cumulus tower

Echo rises at ~400-1700 fpm, and after ~12 minutes tops out at ~ -29°C (31,000

ft

) and descends

Coincident with this is the occurrence of the first lightning stroke (IC)

~ 7 minutes after beginning of the descent, see first CG flashSlide3

Introduction - Development of Initial T-storm

Reynolds and Brook (1956) found six times where precipitation onset was correlated with lightning

Three showed almost simultaneous occurrence

Three showed lightning many minutes after precipitation began

Seems lightning only began with rapid vertical growth in echo (stronger updraft)

Obviously precipitation is a necessary, but not sufficient condition…also need strong updrafts in the cloudSince precip and electrification intimately linked, look at the kind of precipitation that gives way to lightningInitial echo particles 100-300 with a population density of 103-104/m3In semi-arid regions could be ice from the get-go, or quickly freeze (not necessarily the case for midwestern storms)Larger frozen particles form graupel via accretion (with large fall velocities), coexisting with ice crystals growing via sublimationGraupel particles grow in size, more efficiently collect supercooled droplets (sld), s.t. cooling by evaporation/conduction < latent heating of collected sld, glaze ice forms on graupel

 Slide4

Introduction – Charge Structure

Mean temp of negative charge centers involved in IC flashes around -15°C, with limits of -1°C to -33°C

Positive centers usually around 4°C cooler (~2,000 ft. higher)

Other studies agreed with dipole structure, but not mean temps of the centers

Negative charge centers found anywhere between 0°C and -40°C

Also, smaller magnitude/extent pocket of positive charge often found just below negative center (associated with precipitation)Slide5

Electrification Accompanying the Freezing of Dilute Solutions

Workman and Reynolds (1950) found that sign/magnitude of potential difference developed/quantity of charge separated when freezing dilute aqueous solutions functions of kind/amount of contaminant dissolved into the water

Found that water charged positive, while ice charged negatively (based on types of contaminants in NM precipitation)

Reynolds (1953) basically found that glaze ice formation can’t proceed in NM t-storms (with LWC at theoretical max) at T < -10°C

If this is true, then glaze-ice mechanism can’t account for

production of negative charge centers where T < -16°C Slide6

Experiments Under Simulated Thunderstorm Conditions

Reynolds, Brook,

Gourley

(1957)

Apparatus operated in super-cooled cloud,

s.t. speed of the sphere was ~ 25-30 fpsBottom figure shows the chamber in which these exps were operatedIntroduced cloud crystals by seeding w/ solid CO2 fragments or silver iodide smoke, or w/ introduction of rod < -40°CSlide7

Experiments Under Simulated Thunderstorm Conditions

Using framework shown in last slide, found, under certain conditions…

D

uring the riming process large amount of charge separation occurred

Charge separation a result of rubbing contact

Sign of charge separated controlled by ΔT between graupel and ice crystals (and certain contaminants)Above conclusions derived from the following demonstrations…In a cloud with only SLD or ice crystals in equilibrium with water vapor, negligible chargingIn cloud with SLD and ice crystals in coexistence, large amount of charge separatedSlide8

Experiments Under Simulated Thunderstorm Conditions

Attempts were made to establish relative amounts of SLD and ice crystals required for charge separation of a given sign

Varied CWC from 0.25 g/m

3

– 4 g/m

3 and ice crystal concentrations from 104 – 109/m3From ~105/m3 – 108/m3 ice crystal concentration, found lots of liquid water in cloud, and rime formation charged negativelyFrom ~107/m3 – 109/m3 ice crystal concentration, cloud was mostly made of ice crystals, rimer acquired positive charge Rimer collection efficiency is greater for SLD than for ice crystalsRate of collection of rime at a given total CWC an indicator of relative amounts of SLD and ice crystals

Reynolds, Brook,

Gourley

(1957)Slide9

Experiments Under Simulated Thunderstorm Conditions

Found that negatively-charged rimer associated with accretion of large amounts of SLD

In these situations, temp of growing rime ice > temp of ambient cloud

Testing whether a temp difference between rime and ambient cloud could control the charge sign acquired

250-W lamp ~1 ft. from rimer that’d heat the rimer

Set conditions s.t. there’s positive-charging of rimerHeated the lamp, and charge acquired on rimer became negativeTurned the lamp off, and charging on the rimer returned to being positiveAlso found that when contaminants serving as condensation nuclei introduced to cloud, sign of rimer became negative, regardless of SLD or ice crystal concentrationSlide10

Electrification Resulting from Rubbing Contact Between Ice Formations

To perform tests on the rubbing of ice crystals and graupel pellets, used apparatus to right

Cool stationary (probe) rod and vertical (manipulated) rod to -15°C

Dip both rods in water cooled to 0°C

Arranged rods to position in lower-right corner figure

Results found that…Both rods coated with distilled-water ice (clean ice), warmer of the two would charge negatively One of ice formations prepared with 10-4 molar soln NaCl, contaminated ice charged negatively, even with contaminated ice formation being up to 25°C coolerContamination of 2x10-5 molar got similar result as aboveBoth ice formations prepared from solns with concentrations from 2x10-5 to 10-4 molar, little to no charge transfer Ice formations with more concentrated contamination solns do sometimes charge positively Reynolds, Brook, Gourley (1957)Slide11

Application to Thunderstorms

For a normal dipole, need particles with bigger fall speeds to be charged negatively

Experiments conducted show that the precipitation particle must be warmer/more contaminated than particle it rubs against to charge negatively

Use equations from

Ludlam

(1950) to show how this can happen2 mm radius graupel pellet falling at 30 fps in a cloud with a LWC = 1 g/m3, at T = -20°C will be 3.1°C warmer than Tcloud Ice crystals growing via sublimation in presence of SLD at ~-20°C become 0.4°C warmer than TcloudSo, see a double whammy for the graupel to charge negativelySlide12

Application to Thunderstorms

Is this rubbing contact between ice crystals and graupel an appropriate mechanism for thunderstorm

electrification?

50 µm-radius ice crystal and 2 mm-radius riming graupel pellet collision in experiment yielded 5x10

-4

esu/collision With 5x10-4 esu/collision, graupel pellet would gain 5 esu/g(water collected) in a cloud with 1 g(liquid water)/m3 and an ice crystal concentration of 104/m3Most ice crystals likely > 50 µm-radius, making area of contact between the crystal and the graupel pellet greater in a t-storm than in lab settings (therefore, charge per gram collected may be > 5 esu/g)If indeed the charge of the graupel pellets is 5 esu/g, and graupel LWC is 10g/m3, the cell diameter needed to produce a 20 C discharge is 1.3 km Slide13

Application to Thunderstorms

Want to determine the behavior of rainwater in frictional electrification

Used 9 samples from central NM storms in 1954 in t-storm season

Rainwater-ice rubbed against clean ice

 negatively-charged rainwater-ice if as much as 1

°C warmer than the clean ice (in temperature range -10°C to -30°C, this result didn’t vary)Measured resistivities were ~10-4 molar concentration of ions, but five samples acted as though they were clean-ice formations if only 2°C or so colder than the clean-ice formationsFor the other samples, acted as though they had traces of NaCl contamination, and only charged positively if >= 5°C colder Though the type of contamination is important, temperature has a larger modulating effect on what sign of charge a particle will acquire When graupel pellets carried up to T ~ -40°C, find no more SLD, and the temp of graupel and the ice crystals is about the same Rubbing contact statistically symmetric, and with no real temperature difference expected, no charge separation expectedSlide14

Conclusions

Lab experiments show that charge separation is a result of collisions between graupel pellets and ice crystals

In frictional contact, the colder ice formation charges positively, unless it has NaCl concentrations > 10

-5

molar

Main t-storm dipole a result of frictional contact between these two ice hydrometeorsRubbing contact of two ice hydrometeors experiments herein clarify earlier studies of frictional electrification of iceClear from this paper’s studies that temperature and contamination differences between ice formations need to be taken into account when studying frictional electrification of ice