Infrared satellite precipitation estimation algorithms rely on relationships between cloud top properties and rainfall. This study is unique in comparing instantaneous (rather than average) precipitation distributions and lightning strikes to satellite-detected cloud tower attributes. Roughly 2/3 of cloud towers represented areas of strong updrafts rising out of the surrounding cloud mass. Half of the precipitation occurred in the cloud mass outside these towers, matching the probability of finding precipitation in progress when a tower is first detected. Nearly the same pattern occurs for lightning but with lower fractions. Typically, the distribution of instantaneous rainfall rates had the same shape inside and outside the towers, and did not change with tower top temperature or area; yet the frequency of occurrence increased as towers became colder and smaller. This suggests that the location of precipitation events inside a storm does not affect instantaneous rainfall rates, only the frequency of occurrence in time and/or space.
Keywords: satellite, clouds, precipitation, precipitation estimation, clouds and precipitation, satellite precipitation retrieval, infrared satellite algorithms, convective towers, convective precipitation, rainfall distributions, rain rate