Multiple Environmental Influences on the Lightning of Warm-based Deep Convection Simulated with AI

The purpose of the four-year study is to apply our aerosol-cloud model with an electrification component (AC) to assess how the lightning and charge structure of normal warm-based storms are controlled by various environmental factors. These include the instability and aerosol conditions, both for initiation of cloud-droplets and ice crystals.

Focus will be given to reasons for the normal-vs-inverted contrast of charge structure and the land-ocean contrast of lightning.The sequence of objectives is as follows:(1) Upgrade our AC with an AI scheme of deep learning (a 3D ´Convolutional Neural Network´) to accelerate the branching algorithm of the lightning scheme;(2) Improve AC further by extending a lab experiment already set up by applicant Phillips at Lund, characterizing a prolific type of secondary ice production (SIP) pivotal for ice initiation and storm electrification;(3) Validate microphysically and electrically a simulation with AC of two thunderstorms with normal charge structure and a warm base, sampled by aircraft and lightning detection networks in field campaigns in: (i) Oklahoma (MC3E), and (ii) the Amazon rainforest of Brazil (Go-AMAZON);(4) Perform sensitivity tests to quantify the role of lighting in both cases of objective (3), and interpret results with theory.