Center for Climate Research

The rapidly changing climate in recent and coming decades is one of the largest challenges of our time, and deserves immediate attention from science and policymakers. Motivated by this challenge, Dr. Rei Chemke's research focuses on understanding the physical mechanisms underlying human induced large-scale flow changes in the atmosphere and ocean, and their climatic impacts (e.g., changes in precipitation, temperature, ice, etc.). Large-scale climate phenomena play a central role in the distribution of the climate zones on Earth by transferring heat, moisture, and momentum across different regions.

These processes are investigated by applying a theoretical understanding of large-scale thermodynamic and dynamic phenomena to observations and the latest suite of climate models. For example, the two most important uses of climate models are for the attribution of climate change signals to human activity and for multi-decadal climate projections. However, the lack of an observed wind record has prevented scientists over the past two decades from revealing how one of the major components in the climate system, the Hadley cell -- which largely contributes to regional precipitation and temperature patterns in tropical-subtropical areas, has been changing in recent decades, and how much of this change may be attributed to human activity.

To address this knowledge gap, Dr. Chemke derived a mathematical relationship between Hadley cell strength and sea level pressure, and used observations of sea-level pressure to show that the Hadley cell has been weakening in recent decades and that this weakening can be attributed to anthropogenic emissions. The ability of climate models to adequately capture this weakening increases our confidence in their tropical climate projections.

The critical importance of early prediction of severe weather and climate phenomena cannot be overstated, particularly in mitigating the impacts of global climate change. This is especially true in the Eastern Mediterranean and Middle East (EMME) region, identified by the IPCC and WMO as an area highly susceptible to the adverse effects of climate change.

Traditional methods used in operational weather forecasting primarily rely on complex numerical simulations based on physical models. However, these models have their limitations -- they struggle to adapt to new patterns observed in real-time and to effectively utilize the enormous amount of diverse climate data and observations available. While artificial intelligence (AI) has transformed many aspects of human life, it has yet to fulfill its promise in extreme climate event prediction.

Prof. Yinon Rudich’s team is addressing this gap by designing innovative AI models that are both grounded in physics and driven by data. These models are specifically tailored to meet the challenges presented by extreme weather events, focusing on the most pressing climatic issues in the EMME region. Specifically, the group’s studies focus on developing a comprehensive understanding of and predictive capabilities for a variety of critical weather phenomena in the region. These include dust storms from various sources, incidents of high pollution events, heat waves, and floodings. By doing so, the group aims to improve our ability to forecast these events, enhancing our readiness and reducing the potential harm they can cause.

Changes to weather and its extremes in the densely-populated Mediterranean region are key for understanding the environmental implications of climate change, and are of immediate concern to society. In this region, cyclones cause most high-impact weather events, ranging from heavy precipitation, storm surges, strong winds, extreme temperatures and dust storms.

In the lab of Dr. Shira Raveh-Rubin, the atmospheric dynamics and physics governing those weather systems and their impacts are systematically examined, considering thousands of cyclones from the past decades. Understanding these dynamical interactions is key for evaluating numerical model simulations that provide forecasts to reduce risks. This research enables a process-based evaluation of our current modeling capabilities, with the aim to ultimately reduce uncertainties and improve predictions on both weather and climate time scales.