Assaf’s research explores how machine learning can reveal underlying patterns in Earth systems and improve our ability to understand and predict complex environmental phenomena. In one of his current projects, he is exploring Earth’s seasonality through a data-driven lens, aiming to redefine seasonal boundaries based on meteorological factors and evaluating the effect of climate change on the seasonality. In another project, he is quantifying the acoustic pollution generated by human activity and comparing it to natural soundscapes. He is also revisiting the concept of climate change velocity, developing machine learning approaches to measure how quickly climates are shifting across the globe. His previous work has centered on predicting extreme weather events—including wildfires and floods—using explainable machine learning models.