לוח שנה משולב

Some problems in the theory and applications of stochastic processes reduce to solving integral equations with their covariance operators. Usually, such equations do not have explicit solutions, but useful information can still be extracted through asymptotic analysis with respect to relevant parameters. In this talk, I will survey some recent results on such equations for processes related to the fractional Brownian motion: applications include the problem of small deviations, linear filtering, and statistical inference.

Flood is the outcome of complex processes interacting at a range of scales. Flood generation and its magnitude depend on different precipitation and surface properties. As the climate becomes warmer globally, precipitation patterns are changing and, consequently, altering flood regimes. Resolving the expected changes in flood properties requires examining projections of precipitation features most correlated with floods. While the redistribution of mean annual precipitation amounts is generally known, the trends in many other essential factors controlling floods are yet to be resolved. For example, flash flood magnitude is sensitive to space-time rainstorm properties such as areal coverage or storm speed. Still, knowledge of how these properties are affected by global warming is lacking. Maximal rain rates for duration relevant to the watershed’s response time are also crucial parameters controlling the flood discharge. There is some understanding of how extreme rain rates change, but the magnitude and sign depend on the rain duration considered. Changes in frequency and the intra-seasonal distribution of precipitation events also affect flood regimes. Finally, watersheds of different properties are sensitive to different precipitation features, and thus, different watersheds may respond differently to global warming. In this talk, we will present the complexity of flood response under global warming and then focus on two questions: 1) how does global warming affect heavy precipitation events (HPEs) in the eastern Mediterranean, and 2) how these effects are imprinted in the resulting floods in small-medium Mediterranean watersheds.
We simulated 41 eastern Mediterranean HPEs with the high-resolution weather research and forecasting (WRF) model. Each event was simulated twice: under historical conditions and at the end of the 21st-century conditions (RCP8.5 scenario) using the “pseudo global warming” approach. Comparison of precipitation patterns from the paired simulations revealed that heavy precipitation events in our region are expected to become drier and more spatiotemporally concentrated, i.e., we expect higher rain rates on smaller coverage areas and shorter storm durations that, in total, yield lower amounts of rainfall.
These effects have some contradicting signs, and their full hydrological impact on streamflow peak discharge and volume was further explored. Ensembles of spatially-shifted rainfall data from the simulated HPEs were input to a high-resolution distributed hydrological model (GB-HYDRA) representing four small-medium-size watersheds (18–69 km2) in the eastern Mediterranean (Ramot Menashe). Flow volume is significantly reduced in future HPEs, while the change in flood peak is more complicated due to the combined effect of precipitation amount (decreasing) and precipitation rate (increasing). For the watersheds examined in this research, which are mostly agricultural, flood peaks at the watershed outlets are mostly reduced. The dynamics of flood generation at sub-watersheds of different sizes and properties are further examined in this research to understand scenarios for lowering or increasing flood peaks. This study emphasizes that detecting and quantifying global warming impact on space-time precipitation patterns is essential for flood regime projection.

TBA

Learning from data has led to paradigm shifts in a multitude of disciplines, including web, text and image search and generation, speech recognition, as well as bioinformatics. Can machine learning enable similar breakthroughs in understanding (quantum) dynamics of molecules and materials? This is an especially relevant question given that quantum mechanical predictions are now within experimental uncertainties for complex molecular systems (https://www.nature.com/articles/s41586-023-06587-3).

Aiming towards a unified machine learning (ML) model of molecular interactions in chemical space, I will discuss the potential and challenges for using ML techniques in chemistry and physics. ML methods can not only accurately estimate molecular properties of large datasets, but they can also lead to new insights into chemical similarity, aromaticity, reactivity, and molecular dynamics. For example, the combination of reliable molecular data with ML methods has enabled a fully quantitative simulation of protein dynamics in water (https://www.science.org/doi/full/10.1126/sciadv.adn4397). While the potential of machine learning for revealing insights into molecules and materials is high, I will conclude my talk by discussing the many remaining avenues for development.