Satellite observations of terrestrial gamma-ray flashes

Terrestrial gamma-ray flashes (TGFs) are tremendously bright, short bursts of X-rays and gamma rays observed by satellites as they fly over thunderstorms. TGFs have been observed by the Burst and Transient Source Experiment (BATSE) on the Compton Gamma-ray Observatory and the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) satellite. The spectrum of the RHESSI TGFs suggest that they are produced near cloud-top altitudes and are the bremsstrahlung from beams of electrons with energies in excess of 20 MeV. In this dissertation I report on recent analyses of BATSE and RHESSI TGFs.I show that TGFs produce such high peak count rates that they saturate the readout electronics of both RHESSI and BATSE consequently, we do not yet know the maximum brightness of a TGF. I summarize the first RHESSI TGF catalog, which contains data for over 900 TGFs observed since March of 2002 at an average rate of one TGF every 2.35 days. I show that TGFs are not all drawn from the same intrinsic spectrum but that there is no detectable spectral dependance on TGF brightness, geographic latitude, geomagnetic latitude, or the local day or night. Compton reprocessing of hard photons to lower energies in the atmosphere induces a time-delay between the hard peak and the soft tail of a TGF. I show that RHESSI TGFs have a time-delay that is consistent with a source region <21 km, as do the BATSE TGFs once the instrumental dead time is included. The first TGF data from the BATSE Spectroscopy Detectors are also shown to be consistent with a low-altitude source.One of the most exciting results from this work has been the discovery that several BATSE TGFs are actually beams of electrons, rather than photons, hitting the spacecraft. The model for these events suggests that every TGF should also produce a beam of relativistic electrons in the magnetosphere. Because the photons spread out spherically once they are emitted while the electrons stay bound to the local geomagnetic field line, the likelihood of observing electron beams should be roughly 1% of likelihood of observing photon beam TGFs.