Why are Radar and Passive Microwave Profiles Associated with Lightning Probability over Land and Ocean So Profoundly Different?

One of the most striking aspects of the 16-year TRMM dataset is the contrast between continental and oceanic convection, particularly in their vertical structures, microphysics, and severity. It has become well known that oceanic storms have weaker convection, and far lower occurrence and frequency of lightning. Oceanic storms are also well known to have weaker ice scattering signatures and weaker radar profiles, in the sense that higher radar reflectivities do not extend as far into the upper troposphere. From this, a puzzling contrast between land and ocean storms has arisen that has yet to be explained: Given similar radar and passive microwave signatures, land storms have a far higher probability of lightning than the ocean storms that show the same traits. We posit here that either the remote sensing characteristics are not as sensitive to the known microphysics of electrification by non-inductive ice-ice collisions in the presence of supercooled liquid cloud water, or, that the research community has overlooked some property of the hydrometeors detected by these remote sensors that is masking their widely accepted applicability to diagnosing which convective storms should have lightning, or not. Using 16 years of TRMM data from 1997 to 2013, we examine the different radar profiles and ice scattering signatures of precipitation features in archetypes for mid-ocean and mid-continent regions (e.g. central Pacific and Congo). Using hypothesized idealized hydrometeor profiles, and simple radar and passive microwave simulators, we attempt to construct a metric, using measurements available from TRMM to determine the probability of lightning with fewer land-ocean differences than those we have used in the recent past. If successful, this relationship would apply to the radar and microwave data from the Global Precipitation Measurement (GPM) core satellite currently in orbit.