Despite vegetation and soil moisture’s apparent significant role in modulating regional climate variability, the current understanding of these feedbacks originates mainly from highly parameterized climate models and remains insufficiently tested against observations. We have developed advanced statistical methods and applied them to a spectrum of observational data sets, to disentangle the mechanisms underlying the observed land surface feedbacks to the regional climate in North Africa and Australia.

Dust aerosols directly affect human life and play key roles in the Earth's energy and nutrient budgets. A deeper understanding of dust emission, transport, and variability will boost the confidence in predicting dust storms and projecting the Earth's future evolution. We integrate satellite and ground-based observations with process models to expand our knowledge of dust-climate-ecosystem interactions.

Reliable predictions and projections of wildfires are crucial for the development of efficient and effective adaptation and mitigation strategies; however, such projections remain challenging. Part of this challenge comes from the complex interactions between fire, climate, ecosystem, and human. We utilize machine learning techniques, multi-sector observations, and Earth system models to understand the natural and anthropogenic drivers of wildfires, towards the establishment of credible short-term predictions and long-term projections of wildfire regimes.