Estimating Statistics Related to Droplet Collision Using High-Speed Imaging and Off-the-Shelf Image Processing Tools

Abstract

The focus of this study is on the microphysical processes of clouds that are essential for improving the accuracy of climate models, particularly the analysis of droplet collisions and coalescence using high-speed imaging and off-the-shelf image processing tools. The temporal dynamics of droplet collision events are captured using the implemented processing package, and the average collision probability and coalescence efficiency for the observed droplet size range are calculated as statistical measures related to droplet collision-coalescence efficiency. By combining publicly available tools, a processing procedure with high reproducibility and implementability is achieved. Furthermore, the collision detection conditions specific to droplet collision extraction are defined, the patterns of false positive samples in the extraction results are classified, and their challenges are identified. The statistical measures calculated from the colliding droplets are consistent with previous studies, confirming the validity of the approach. This study proposes a fundamental analysis method for future observations of droplet collisions in turbulent fields using three-dimensional imaging methods such as inline holography.