WESTERN REGION TECHNICAL ATTACHMENT
NO. 97-20
JUNE 17, 1997

NATIONAL LIGHTNING DATA ON THE
WESTERN REGION WIDE AREA NETWORK

Andy Edman - WRH/SSD - Salt Lake City, UT

Background

For the last 15 years, Western Region (WR) NWS offices have acquired lightning data through a cooperative agreement with the Bureau of Land Management (BLM). The BLM established a lightning network over the western United States to support a multi-agency fire suppression program. WR processed BLM lightning data on an AFOS era processor located at the Boise Forecast Office. The Boise computer generated a series of lightning AFOS graphics and special alerts that were distributed to WR offices via the AFOS communication network. This was an effective system and served WR forecasters well. A combination of two events, the failure of critical non-replaceable components of the Boise computer coupled with BLM decommissioning the lightning network during the spring of 1997, brought an end to access of this data set.
To replace this loss of operational data, WR/SSD developed a solution to acquire lightning data from the National Lightning Detection Network (NLDN). The NLDN is a national system deployed and operated by the Global Atmospherics, Inc. (GAI) company. The NLDN is the only national lightning network data available at this time. Since GAI is owned by the Sankosha Corporation, the NLDN is a private sector data set. As such, the NWS must pay a yearly site license fee to use the data and offices cannot share data with any other government or private sector entity. National Weather Service Headquarters plans to include NLDN data as part of the AWIPS data stream.
GAI has incorporated numerous technical advancements to the lightning detection sensors and network. Many of these advancements should aid WR offices. This Technical Attachment will summarize how this network operates and describes how the data are available to the WR Forecast Offices.
Lightning Basics

A single Cloud to Ground (CG) lightning flash is composed of several components. Initially, a weakly charged, normally invisible to the human eye, collection of electrons called Step Leaders begin to move toward the ground from the cloud's base. Each Step Leader advances approximately 50 to 100 meters and last approximately 1 microsecond. A series of Step Leaders is required to move downward through the sub-cloud environment. As the Step Leader approaches the ground, a positively charged Ground Streamer advances from the ground to meet the Step Leader. Once they meet, an ionized channel is formed and electrons can move quickly toward the Earth's surface. As electrons move downward, the electrons recombine with the positively charged ions, first near the surface then moving upward toward the cloud. This recombination produces the brilliant flash, the rapid heating of the air near the ionized channel (thunder) and is called the Return Stroke. Each return stroke exhibits a peak current of 5 to 300 kiloamps and has a nominal duration of 20-50 microseconds. A single flash is normally composed of 2 to 3 return strokes and but can vary from 1 to 20 strokes. A single flash can also contain return strokes that do not follow the same ionized channel. In these cases, the lightning flash can appear as a series of return strokes that are within a few hundred meters of the original channel. In general, most observing systems treat these cases as a single flash. Any good basic meteorology text can provide a more thorough description of the lightning process.
Each cloud-to-ground (CG) return stroke also produces a unique electromagnetic pulse or signature that can travel through the atmosphere for hundreds of miles. It is this electromagnetic signature that most conventional surfaced based lightning detection systems observe. Thunderstorms also produce a variety of other lightning flashes (i.e., cloud-to-cloud, intra-cloud, etc.). Since CG flashes produce the most impact to the public and a consistent electromagnetic signature, conventional lightning data observing systems screen out most of the non-CG lightning flashes. The NLDN provides information about a CG lightning flash which is composed of multiple CG return strokes. The WR system displays CG lightning flashes. The number of return strokes for each flash is part of the NLDN data, if desired.
How the NLDN Operates

For the last 20 years, ground based CG lightning observing systems used one of two basic technologies, Time of Arrival (TOA) or Direction Finding (DF). TOA systems work by listening for the electromagnetic pulse and recording the precise time it was received. This information is then relayed to a central processing site, where the data from a number of sensors are combined and through the use of spherical geometry, a location solution is computed. In order to compute an accurate flash location, three or more sensors are required to observe the lightning flash. Also, maintaining the exact time among a set of widely dispersed sensors was critical to the accuracy computations. TOA systems were also marketed as LPATS systems and were used in the Midwest and southeast United States. The DF class of lightning observing systems used a number of sensors that recorded the time of the pulse and the direction from which the pulse originated. The data was sent to central processing unit, where input from a number of sensors was used to triangulate a solution. Exact timing was not quite so critical with this system, but the accuracy of the angle measurement could cause considerable error. This technology was also known as LLP and was used by SUNYA, BLM, and the NSSL networks.
GAI acquired the patents for both systems, combined the technologies, and added sensor enhancements to improve the system. The new sensors, called Improved Performance from Combined Technology (IMPACT), use the direction finding capability to provide the azimuth (direction) information while the TOA capability is used to accurately assign a time to the electromagnetic pulse. This allows flash solutions to be formed from only two sensors in sensor poor areas. It also allows for additional mathematical optimization to be used to further refine the final flash locations and time. Wave form discrimination was also installed on the sensor to reduce spurious noise and is currently used to filter out non-CG flashes. The result is that the NLDN data should be better than either of its predecessors.
GAI currently uses a mixture of IMPACT and upgraded LPATS sensors that contain much of the IMPACT capabilities. Figure 1 shows the location of the sensor network. Data from each network is uplinked to a central processing unit called the Network Control Center (NCC) located at Tucson, Arizona through a satellite communications system. The NCC computes the location of the lightning flash in real time and uplinks the processed lightning location data through the same satellite system to their customers. Information about a lightning flash is available typically within 40 seconds of its occurrence.
Accuracy and Detection Efficiency

Historically, the observed skill of a lightning network has been measured by the accuracy of locating the lightning flash and how many lightning flashes were observed out of the total produced by a storm (detection efficiency). These measurements are tough to validate and often have been the source of considerable scientific debate. Figure 2 is the estimated accuracy, and Fig. 3 is the estimated detection efficiency for the NLDN network as provided by GAI. In general over WR, detection efficiency is around 80 percent with .5 to 1 km location accuracy. Please note how both detection efficiency and location accuracy fall off rapidly near the coasts and U.S. borders. This is a result of the network configuration. What is of more interest is the improvement made during the last five years by combining the two technologies. Detection efficiency has improved by 10-20 percent and location accuracy has improved by 4-8 km over the old LLP system.
How the WR System Works

The old WR Boise system transmitted graphics. The new system transmits digital data about each lightning flash directly to each office. National Weather Service Headquarters acquires lightning data from GAI downlink at NWS Telecommunication Gateway. The data are grouped into communication packets and are uplinked over the AWIPS Satellite Broadcast Network (SBN) to any NWS office with a AWIPS satellite antennae. In WR, the Salt Lake Forecast Office has an early AWIPS system. The data are then re-transmitted through the internal WR Wide Area Network (WAN) to each office that has a site license to use the data. The data can then be displayed on RAMSDIS and the Unix Workstations. On RAMSDIS, the lightning flash data can be animated over the imagery to provide a near real-time depiction on how the lightning activity has been changing during the last 30-45 minutes. Due to limitations in the current WR WAN bandwidth, the lightning data are updated every 7 minutes. A special thanks to Kevin Schrab (WR/SSD) and Dave Tomalak (NWSFO Great Falls) for setting up this system.
Summary

Lightning data are a vital operational data set for Western Region offices. While the impact of lightning data is well known for support of the fire weather program, offices also use lightning data in combination with satellite data to supplement the warning program in areas where the WSR-88D coverage is blocked by terrain or the radar beam is above the convection. In addition, the impact of lightning strikes on the general public is growing. I have attached a paper written by Ron Holle (et al.) on the growing number of damage and insurance claims caused by lightning. Ron collected data for three states and extrapolating for similar damage over all 50 states. Insurance claims, resulting from lightning damage, exceed 300 million dollars a year. While damage statistics vary greatly from year to year, damage from lightning is now approaching damages incurred by hail and straight line thunderstorm winds. Over the next several months, additional Technical attachments will be issued highlighting other lightning issues.
References

Holle, Ron (et al.). 1996: Insured Lightning Caused Property Damage in Three Western States, J. Appl. Meteor., 35, No. 8, pp 1344-1351. (Attached to this TA - Hard copy only)
McCollum, Darren, D. Bright, J. Meyer, and J. Glueck. 1996: Operational Applications of the Real-time National Lightning Detection Network Data at the NWSO Tucson, AZ. WR Technical Memorandum 241.
NLDN: A Combined TAO/MDF Technology Upgrade of the U.S. National Lightning Detection Network, Kenneth Cummins, et al., January 1996, 12th International AMS Conference on Interactive Information and Processing Systems (IPPS). pp. 347- 355.