This paper explains a method of locating groups of satellite-observed VHF s
ignals from lightning that emanate from isolated storm regions. LF/VLF sign
als recorded and located by the National Lightning Detection Network (NLDN)
are often accompanied by broadband VHF emissions of sufficient intensity t
o trigger satellite-based RF receivers. It has previously been shown that e
vents recorded by the FORTE satellite and the NLDN, which have occurrence-t
ime differences of 0.3 ms or less, can be assumed to originate at the same
approximate geolocation with 97.5% reliability. For such event pairs, the V
HF events recorded by FORTE were assigned the NLDN locations. We use graphi
cal analysis to identify groups of satellite data that can be assumed to or
iginate from the same storms as the located events. To accomplish this, a v
alue of total electron content (TEC) is derived from the frequency dispersi
on of each VHF signal recorded by FORTE. Scatterplots of TEC versus time of
ten reveal curved clusters of satellite data. The variation of TEC with sat
ellite elevation angle, and time, is consistent with a model in which the l
ength of a signal's required propagation path through the ionosphere change
s with satellite and storm position. We use the values of TEC for the locat
ed satellite data, a simple model of the ionosphere, and NLDN data to help
us identify TEC clusters that can be assumed to originate from relatively c
ompact, and unambiguous, geolocations. For the time period of April through
September of 1998 we geolocate 65 groups of satellite data records. A tota
l of 6131 satellite data records, which were not previously located, have b
een assigned the median NLDN coordinates of their corresponding storm.