A method of preprocessing lightning-measurement waveforms has been devised to reduce the bandwidth needed to transmit data for computing the locations of lightning strikes. The method is used in a system in which electric fields and electric-field derivatives induced by lightning are measured at remote stations and the measurement data are transmitted to a central station. At the central station, the location of each lightning strike is computed from the known positions of the stations, the speed of light, and the differences among times of arrival of common waveform features at the remote stations. Prior to the development of the present method, in order to achieve a required spatial resolution of tens of meters, it was necessary to transmit full measurement data from each remote station to the central station in real time at a bandwidth of 10 MHz.
The preprocessing method exploits the following two concepts:
- A lightning strike typically comprises multiple strokes 20 to 50 ms apart, each stroke lasting of the order of 0.1 ms. Even during an intense thunderstorm, lightning strikes usually occur no more frequently than once every few seconds. Given this essential intermittency, the data-transmission bandwidth necessary to accommodate the time-averaged data rate is much less than the real-time, full-resolution bandwidth of 10 MHz. Thus, by abandoning real-time transmission, one can make possible a reduction in the data-transmission rate.
- If one extracts data on waveform characteristics that can be used to establish times of arrival, then the remaining waveform data can be discarded to effect a further reduction in bandwidth.
In this method, each remote station is equipped with a digitizer and an embedded controller. When a signal exceeds a predefined threshold, digital samples of the signal are taken at a rate of 20 megasamples per second for an observation interval of 100 µs. The record of samples is time-stamped with the starting time of the observation interval as determined by a differential Global Positioning System (GPS) receiver.
Still at each remote station, waveforms are characterized in terms of time elapsed between largest peaks (both positive and negative), rise times, durations at half maximum amplitudes, and other parameters. Each such characteristic is encoded in two bytes: one that identifies the characteristic and one that gives its value. The resulting code data and the time stamp are then transmitted to the central station for processing.
In the central station, the data from the remote stations are lined up in a search for matches among the characteristics. Differences among times of arrival are then determined on the basis of the best match.
The accuracy of the estimated differences among times of arrival increases with the number of characteristics. About 10 unique characteristics are enough to obtain good alignment. The bandwidth needed to transmit 10 coded characteristics and the accompanying timing data is 5 kHz — only 1/2,000th of the bandwidth needed for real-time transmission of full waveform data.
This work was done by Jose M. Perotti of Kennedy Space Center and Pedro J. Medelius, formerly of I-NET, Inc. For further information, access the Technical Sup-port Package (TSP) free on-line at www.nasatech.com/tsp under the Electronics & Computers category.
Inquiries concerning rights for the commercial use of this invention should be addressed to
the Technology Programs and Commercialization Office
Kennedy Space Center
Refer to KSC-11955.