HAiVA: Hybrid AI-assisted Visual Analysis Framework to Study the Effects of Cloud Properties on Climate Patterns
This work addresses the need for computationally efficient tools for climate scientists to study cloud intervention scenarios, though it appears incremental as it builds on existing AI and visualization methods.
The authors tackled the problem of understanding cloud-climate interactions for climate intervention techniques like Marine Cloud Brightening by proposing a hybrid AI-assisted visual analysis framework, which facilitates interactive what-if investigations to assess impacts on climate patterns.
Clouds have a significant impact on the Earth's climate system. They play a vital role in modulating Earth's radiation budget and driving regional changes in temperature and precipitation. This makes clouds ideal for climate intervention techniques like Marine Cloud Brightening (MCB) which refers to modification in cloud reflectivity, thereby cooling the surrounding region. However, to avoid unintended effects of MCB, we need a better understanding of the complex cloud to climate response function. Designing and testing such interventions scenarios with conventional Earth System Models is computationally expensive. Therefore, we propose a hybrid AI-assisted visual analysis framework to drive such scientific studies and facilitate interactive what-if investigation of different MCB intervention scenarios to assess their intended and unintended impacts on climate patterns. We work with a team of climate scientists to develop a suite of hybrid AI models emulating cloud-climate response function and design a tightly coupled frontend interactive visual analysis system to perform different MCB intervention experiments.