Lightning Mapping Arrays: Technical Status and Developments

Lightning Mapping Arrays:
Technical Status and Developments
William Rison, Paul Krehbiel, Ron Thomas,
Graydon Aulich, Harald Edens
Langmuir Laboratory for Atmospheric Research
New Mexico Institute of Mining and Technology
Socorro, NM 87801
Southern Thunder 2009
Cocoa Beach, Florida July 28, 2009

LMA Operation
●
Listen in a locally unused TV channel
●
Detect peak event in successive
80 microsecond time intervals
●
Measure arrival time within 40 ns
●
Up to 12,500 arrival times/second

Height vs. N-S
Example of a highly dendritic negative cloud-to-ground (CG) flash
Height vs. time
Height vs. E-W
60 km extent,
2.5 sec duration,
7350 sources
Plan view

Oklahoma Lightning Mapping Array (OU, NSSL)
●
Density of points display
Kansas
Oklahoma
Texas
(http://lightning.nmt.edu/oklma)

VHF Time-Of-Arrival (TOA) Total Lightning Mapping Systems
LMA
LDAR-II
DPG
(?)
LLAB(3)
TSC
(?)
(N) VHF TV Channel
WSMR
(3)
STEPS
2000 (3)
DFW
(3)
OK(3)
HOUST
(~4)
HSV
(5)
ATL
(9)
DC
(10)
KSC
(3-4)

Practical Considerations
●
VHF Frequency: Lightning measurements are best made in the lower VHF
– Radiated source power decreases as ~1/f 2
– Antenna gain decreases as 1/f 2 [(1/2)-wave dipole antennas]
– Decreased detectability, range in upper VHF
●
Array Size
– Station spacing of ~20 km necessary for 3D accuracy and good
sensitivity
●
Number of Stations
●
– Better accuracy and detectability with more stations
– Minimum number: 10
●
Communications Link
– Need 150 Kb/sec for decimated real-time data
– Need 1 Mb/sec for full data
– 802.11 Wireless (OKLMA, NALMA)
– Internet (DC)
– Fiber Optic (WSMR)

VHF Frequency Spectrum
LMA stations ‘listen’ on a locally unused TV channel (e.g., Ch. 3)

Total Lightning Observations

Oklahoma LMA
●
Eleven-station network,
60 km diameter, southwest
of Oklahoma City.
●
Central processing
station in Norman
● 802.11b wireless
communications

Oklahoma Lightning Mapping Station

Oklahoma LMA
●
For 11 station network,:
Good 3-D to about 150 km
Good 2-D to about 250 km

Oklahoma LMA
●
Inside 3-D region,
significant detail:
●
– Upper positive,
– Mid-level negative,
– Lower positive
– Good sensitivity
●
Much less detail in 2-D
region
– No sources below ~7 km
– Little altitude information
– Low sensitivity, only
see higher-power sources

VHF Time-Of-Arrival (TOA) Total Lightning Mapping Systems
LMA
LDAR-II
DPG
(?)
LLAB(3)
TSC
(?)
(N) VHF TV Channel
WSMR
(3)
STEPS
2000 (3)
DFW
(3)
OK(3)
HOUST
(~4)
HSV
(5)
ATL
(9)
DC
(10)
KSC
(3-4)

Real Time Processing and Display
●
Full Data: typically one trigger every 80 µs
●
Real-time Decimated data: one trigger every 400 µs
– Data links typically not fast enough
– Cannot process 12-station real-time data in real time
●
For an active storm, number of decimated sources about one third the
total of the full data set
●
A few second to a few minute delay in source location generation

●
Decimated data or full data?
●
Number of active LMA stations
– Significant reduction in number of located sources with fewer than
eight active stations
●
VHF frequency used
●
Local noise threshold
●
Proximity to array center
Interpretation of Source Density
●
WSMR LMA most sensitive system
●
DC LMA least sensitive system

Volcanic Lightning
Augustine Volcano
Nov. 4., 2006 Photo by Jennifer Adleman, courtesy of AVO/USGS

Portable LMA Stations
• Electronics housed in shielded
thermoelectric cooler enclosure
• Operate from external 12 VDC battery and/or
power supply. ~12 (+48) watts power
• Battery operation: 48+ hours (w/out cooling)
20+ hours (with cooling)
• Lightweight (10 lbs)

LMA Antenna –
AVO Facility,
Homer, Alaska
(Augustine in the
distant background,
hidden by clouds)

Received signal at Homer during the main eruption on January 28
Two types of electrical activity
-30
Continuous sparking
Discrete (< 1 sec)
-50
Power(dBm)
-70
05:20 05:30 05:40 05:50
Augustine Seismic
Oil Point Seismic

Expected arrival time differences.
Two stations give the azimuthal direction only

700 ms horizontally extensive lightning discharge
in Augustine volcanic plume

40 ms vertical lightning discharge in Augustine volcanic plume

Three Types of Electrical Discharges
-30
-50
Discrete short
duration (< 100 ms)
Discrete long
duration (< 1 sec)
Power(dBm)
-70
Continuous Sparking
Thee types of electrical activity:
05:25 05:30 05:35
1) Continuous sparking, presumably in hot gases at the mouth of the crater
2) Short-duration discharges, presumably negative leaders ascending into positive
charge during rapid vertical development of the volcanic plume
3) Longer duration lightning discharges, with both vertical and horizontal
development, in the anvil of the volcanic plume

Redoubt Volcano Eruptions, Alaska, March 22 – April 4, 2009
22 eruptions; 4-station Lightning Mapping Network ( )

March 28, 2009 eruption
●
~30 min duration (2330 --
0000 UTC), major
lightning.
●
Electrical activity began
over Redoubt, but became
detached from volcano and
drifted downwind with time.
●
Cloud was self-electrifying
after becoming detached.
●
Charge structure of final
discharge similar to upper
dipole of normally-electrified
storm.

Volcanic Lightning
●
Lightning is common in volcanic eruptions
●
Horizontally extensive lightning in ash cloud
●
Not practical to routinely map lightning using VHF TOA system
●
Geostationary lightning mapping sensor might provide valuable
information on volcanic ash clouds

Comparison of NEXRAD and Lightning Source Density
(Oklahoma LMA)
Radar
LMA (plan view)
LMA (vertical cross-section)
Courtesy of Lakshmanan, Hondl, MacGorman, Stumpf (CIMMS, 2004)

Washington, DC Greater Metro Area Lightning Mapping Stations
(8 Stations, 2 more stations to be added, Spring 2008)
APL
UMBC
Sterling
MCCC
Howard
GMU
NVCC
CSM
Joint Project: New Mexico Tech, NASA Global Hydrology and Climatology Center,
NOAA National Weather Service

DC Lightning Mapping Array
●
Experimental, demonstration network
-- Urban environment, operates in upper VHF (Ch 10)
-- Use internet for data communications (real-time, post-storm)
●
Real time data processing done by NASA and provided to NWS
for nowcasting the weather.
●
Real time data at: http://branch.nsstc.nasa.gov/PUBLIC/DCLMA
Hosting Institutions:
●
Howard University
●
Montgomery County Community College – Rockville, MD
●
NWS Test and Evaluation Facility – Sterling, VA
●
College of Southern Maryland – La Plata, MD
●
Johns Hopkins – Applied Physics Laboratory
●
North Virginia Community College – Annandale, VA
●
University of Maryland Baltimore County (UMBC)
●
George Mason University (Prince William Campus, Manassas).
8 station network, 2 more stations to be added, Spring 2008

Isolated storm, Washington DC
August 3, 2006
Sterling
Rockville
UMBC
Howard
●
1 hour of data; storm propagates
west to east, from Sterling, VA
to over DC area (40 km)
●
Note vertical growth phases in
height-time panel (top).
●
Indicates growth of successive
cells in the storm (plan view)
●
Storm produces a number of
‘bolt-from-the-blue’ discharges
out the sides and ahead of the
storm (‘spider’-like channels).
Annandale
La Plata

Bolt from the Blue Discharge
© Harald Edens

Real-time public website ( http://branch.nsstc.nasa.gov/PUBLIC/DCLMA/)

Sept 8, 2006

October 4, 2006

DC LMA diagnostic webpage
Data ingest time (seconds)

April 4, 2007
●
Two hour time interval
●
Active Upper and Lower
Positive Charge Regions
=> Normal tripole
charge structure

April 4, 2007
●
10 min time interval
●
Density of points plot

April 4, 2007
+ --
+ -
●
Charge density plot
●
(+) charge regions (orange)
better detected than (-)
charge regions (blue)

Intracloud Discharge
(+)
(--)
●
Between negative (blue)
and upper positive (red)
charges

CG Discharge
●
Between negative (blue)
and ground, via lower
positive charge (red)

Low IC Discharges
●
Between negative (blue)
lower positive charge (red)
Low Intracloud

Summary:
+
--
+
●
High density of points
=> (+) charge regions
●
(Negative in between)

May 13, 2007
●
Active lower (+) region
(Low ICs and -CGs)
●
Upper (+) becomes active
during convective growth
phases.

SCAN Storm Cells / Site Storm Threat super imposed with DC LMA image (July 04, 2007 at
23:01Z
Red > 60
Red > 6
Yellow: 2-6
White : 1-2
Gray < 1
SCAN CELL TABLE
Cell S1
DC LMA total lightning
Red > 6
Yellow: 2-6

AWIPS Display, NWS Forecast Office, Huntsville, AL
VIL
LMA
2-min
Composite
dBZ
NLDN
5-min

Bolt from the blue
lightning discharge
(negative cloud-ground)

Total Lightning Observations

Total Lightning Observations

May 27, 2007

May 27, 2007

February 22, 2007
Winter storm, low lightning
rates, lower altitudes

April 4, 2007
Active Upper and
Lower Positive Charge
Regions

REFERENCE:
Thomas, R.J., P.R. Krehbiel, W. Rison, S.J. Hunyady, W.P. Winn, T. Hamlin,
and J. Harlin, Accuracy of the lightning mapping array, J. Geophys. Rsch.,
109, D14207, doi:10.1029/2004/JD004549, 2004.

New Mexico LMA Networks

WSMR – Aircraft sparking observations
Aircraft becomes electrically
charged in flying through
ice crystal clouds (cirrus,
thunderstorm anvils). Emits
steady stream of weak sparks
that can be located by the
LMA. 100 to 10,000 sparks
per second, depending on
charging rate.

L a ng m uir L a bora to ry
C o m pa c t L MA

Total Lightning Observations
●
Clouds are transparent at RF – therefore can use to sense intracloud (IC)
discharges as well as in-cloud parts of cloud-to-ground (CG) lightning.
=> Hence, “total” lightning.
●
Lightning is a good indicator of vigorous convective activity --
valuably complements radar observations of storms by showing
most active and intensifying parts of a storm system.
●
Total lightning observations now being made at several locations
around U.S.
●
Data starting to be used by National Weather Service to nowcast
potentially severe weather situations, lightning ground hazards.
●
Potential benefit to air traffic control, airport operations.
●
VHF observation technique potentially useful in other applications
-- commercial, defense-related.

Multiple storm observations – June 17, 2006
●
830,000 sources in 1 hr;
(200-250 per sec)
●
Increasing minimum height
in vertical projections
due to earth’s curvature
(line of sight propagation)
●
Difference from radar:
Essentially instantaneous
pictures of storm activity
(no need to scan).

Tornadic Storm, June 29, STEPS 2000: Height vs. time density plot
convective surges
F1 tornado
F1 tornado preceded by 2 convective surges (A, B) 45 min earlier;
accompanied by 3 rd surge and by onset of +CG discharges

Lightning ‘hole’ associated with F1 tornado
●
Tornado on west side of hole
●
Associated with 3 rd convective
surge of storm – note upward surge
in height-time panel (top).
●
Hole co-located with strong
(100+ mph) updraft

DC LMA Stations (Urban Network)
Howard University
Montgomery Comm. College, Rockville
North VA Comm College,
Annandale
Sterling Nat'l Weather Service Site

Airplane Track
Aircraft becomes electrically
charged in flying through
ice crystal clouds (cirrus,
thunderstorm anvils). Emits
steady stream of weak sparks
that can be located by the
LMA.

DC LMA diagnostic web table
December 12, 2006; quiescent

Examples of 24-hour lightning activity, DC mapping array
Oct 4, 2006 Sept 8, 2006
(http://branch.nsstc.nasa.gov/public/dclma)

5 min intvl

LMA Operation
●
Simple receiver
●
6 MHz bandwidth, t_rise ~160 ns
●
Detect peak event in successive
80 microsecond time intervals
●
Measure arrival time within 40 ns
(25 MHz A/D converter)
●
Up to 12,500 arrival times / second

Oklahoma Lightning Mapping Array (Univ. of Okla., National Severe Storms Lab)
Kansas
Oklahoma
●
One hour of data.
●
Eleven-station network,
60 km diameter, southwest
of Oklahoma City.
●
Density of points display
●
Squall line approaching
Oklahoma City from west.
●
Operates on TV Ch. 3
(60-66 Mhz, lower VHF)
●
Array covers most of OK
and into KS and TX.
(400-500 km diameter area)
●
Data ingested over wireless
ethernet links, processed in
real time; 1-2 min updates.
Texas
(http://lightning.nmt.edu/oklma)

DC LMA diagnostic web table
August 3, 2007 storm, in progress

Concluding Remarks
●
Very helpful to have locally unused TV channels in the VHF.
6 Mhz bandwidth provides good timing accuracy, sensitivity.
●
Lightning measurements are best made in the lower VHF
-- Radiated source power decreases as ~1/f^2
-- Antenna gain decreases as 1/f^2 [(1/2)-wave dipole antennas]
●
Lower versus upper VHF (Ch 2-6 vs. Ch. 7-13):
-- factor of 2-3 difference in frequency
-- (6-10 dB)*2 = 12-20 dB decrease in sensitivity
-- decreased detectability, range in upper VHF
●
Urban environments: Stronger local noise in lower VHF, can
benefit from operating in upper VHF.
●
Lightning data will be increasingly useful to the NWS and other
governmental departments and agencies (public safety, national security).
●
If at all possible, leave a lower VHF channel free for
this and future remote sensing applications

Intracloud Lightning

Cloud-t o-Ground
Light ning