An exploration framework to identify and track movement of cloud systems

Area of Science:

• Atmospheric Science
• Computational Topology
• Computer Vision

Background:

• Cloud systems exhibit complex movements across diverse spatial and temporal scales.
• Phenomena like the Madden Julian Oscillation (MJO) involve large-scale oscillations with embedded smaller cloud clusters.
• Understanding these multi-scale interactions is crucial for weather and climate studies.

Purpose of the Study:

• To develop and present a novel framework for exploring and visualizing cloud system dynamics.
• To enable the study of cloud movements at multiple scales without user-defined prior knowledge.
• To analyze multi-scale interactions within tropical weather systems.

Main Methods:

• Utilizing computational topology and computer vision techniques.
• Developing an automated framework for delineating cloud clusters.
• Applying the framework to analyze large-scale oscillations and embedded smaller-scale systems.

Main Results:

• The framework successfully delineates cloud clusters automatically, independent of user expertise.
• Analysis reveals multi-scale interactions within tropical systems, including cyclones.
• Demonstrated effectiveness in tracking organized cloud systems during a Mumbai rainfall event and Cyclone Aila.

Conclusions:

• The developed framework provides a powerful tool for studying cloud system dynamics.
• It facilitates the discovery of previously speculative multi-scale interactions.
• The approach is effective for analyzing complex weather events and tropical cyclones.