Why does cloud-ground lightning occur so much less frequently over ocean? I was talking with an acquaintance about lightning, and we came up with opposite theories and predictions for the frequency of lightning over ocean.


*

*My theory is that since seawater is a fluid riddled with ions, the charges can move to equalize with whatever field is applied to it; therefore it should take a much higher, much faster increase in the magnitude of cloud charge to cause the dielectric breakdown necessary for a strike--in other words, the clouds have to beat the ocean to the punch before it dissipates the increase. So I predict less frequent lightning over the ocean.

*His theory is that the composition of the lower surface is irrelevant; it's the insulating effect of the air that contributes to favorable conditions for lightning generation. He predicts more or less equal frequency of lightning over land or sea.
Wikipedia bears out my prediction (sourced from NASA data):

According to the distribution of lightning page, strikes occur most frequently in the tropics, on land. What makes this possible? More importantly, which of us is right? ;)
 A: This is not my area but I found an interesting article about lightning on the NASA web site. A key point from this article is that only 25% of lightning is cloud to surface and the rest is made up from strikes between clouds. Therefore changes in the surface or the insulating properties of the air near the surface could affect only 25% of the lightning strikes. But much larger differences are seen between land and sea, and this suggests that the explanation lies elsewhere.
I would guess the reason is that the special type of weather systems needed to generate lightning do not occur over the oceans and that is the reason for the dearth of lightning. The NASA article makes the point that lightning is most intense where air masses converge e.g. Florida and the Himalayas. Over the oceans where there are no topographic features to deflect the wind you would not get the air mass convergence that is required to generate lightning.
A: I want to note a slight modification or correction to John's answer.  Hurricanes/typhoons/cyclones etc. all generate a tremendous amount of lightning and numerous other storms found over the oceans do so as well.  This is somewhat indicated by the image shown which has enhanced lightning strikes in regions where these storms form or are common.  So the argument that lightning is less frequent over the oceans is not really sound, I do not think.
The biggest difference between the ocean and the Earth's surface is the conductivity.  The conductivity of ocean water can be as high as ~5-6 S/m [e.g., Tyler et al., 2017] while sand and silt tend to be >10-100 times lower.  Higher conductivity materials accumulate charge less than lower conductivity materials, i.e., electric fields need to do less work to redistribute charge in higher conductivity materials and net charge accumulation is lost in fully conductive materials.  For comparison, copper has a conductivity of more than 10 million S/m, i.e., ~2 million times higher than ocean water.
In summary, I think the biggest issue is that soil is more likely to accumulate a charge imbalance with a cloud allowing large electric potentials to build up until an electrostatic discharge is possible.  Ocean water can polarize under a cloud which means charge build up within the cloud needs to be stronger and faster in order to initiate a discharge in the atmosphere to the ground.
