There is clear evidence that a catastrophic resurfacing event occurred on V
enus in the relatively recent past. A primary data source for this resurfac
ing event is the spatial distribution of impact craters. In this paper, we
apply the pair-correlation technique to the observed crater distribution an
d find that the result is identical to that for a random distribution. In o
rder to test the sensitivity of the technique, we also apply it to the spat
ial distribution of coronae on Venus. For the coronae, we find substantial
deviations from a random distribution. One explanation for the catastrophic
resurfacing is the episodic subduction hypothesis. We model episodic subdu
ction using a thermal boundary-layer stability analysis. We find that episo
dic subduction events with intervals of 500 to 700 Myr can transport only 1
5-25% of the radiogenic heat produced within the planet. We suggest that th
e remainder of the heat must be lost to the surface during a period of vigo
rous tectonic activity, following the subduction event but prior to the sub
sequent stabilization of a global lithosphere. (C) 1999 Academic Press.