Summary and Conclusions
We have described the photodiode detector (PDD) on board the FORTE satellite and presented examples of optical measurements of terrestrial lightning obtained from the PDD. The PDD event data set has been compared with data obtained with the National Lightning Detection Network (NLDN). An example of dual PDD/NLDN stroke detections was presented, and the NLDN-reported cloud-to-ground (CG) stroke detection efficiency of the PDD was found to be 
6%. Those PDD events that are associated with NLDN-reported CG strokes represent only 8% of the PDD event data set. Furthermore, there is evidence that the PDD detection efficiency is significantly higher for events occurring in maritime cloud conditions compared to continental cloud conditions. On the basis of the overlap between the NLDN CG stroke data and the PDD event data we conclude that the PDD preferentially detect the in-cloud components of total lightning.
The PDD event data compare favorably with previous satellite- and aircraft-based measurements of optical lightning emissions. The median peak power observed by the PDD is nearly 1 
10 9 W, consistent with observations made by Turman [1977] using data from a DMSP satellite. The median optical energy of source events is estimated to be 450 kJ, which is consistent with previous ground-based and aircraft-based measurements. Both the peak optical power and optical energy estimates for the sources represent lower bounds since off-nadir and atmospheric extinction factors were neglected in this work.
The effective pulse widths of PDD events are much larger than those observed in ground-based measurements but are consistent with aircraft-based measurements. We find the median effective pulse width to be 590 
s for a general population of PDD events. The median pulse width is slightly larger for those PDD events associated with NLDN-reported CG strokes. We found that in general, the effective pulse width varies inversely with peak optical power. We speculate that this relationship may result from varying amounts of cloud-induced extinction and signal broadening. Direct line-of-sight observations of the lightning channel with small amounts of intervening clouds are expected to result in brighter, narrower pulses than observations with large amounts of intervening clouds between the source and the sensor.
We conclude that the PDD does indeed detect and accurately characterize terrestrial lightning. However, the detected lightning signal typically originates from within thunderclouds, rather than from below the cloud. Thus PDD data are complementary with ground-based measurements.
Next: Acknowledgments Up: Optical observations Previous: Statistical Characte
© 2001 American Geophysical Union