Integrated Calendar

Thick halite sequences are common in the Earth’s geologic record; they were accumulated in deep perennial hypersaline water bodies, saturated to halite and subjected to negative water balance. For decades, evaporites research gained insights from exploring modern shallow hypersaline environments, including the relations between the hydroclimatic forcing and the deposited halite layers. However, there is a knowledge gap in understanding limnological controls on accreted halite sequences in deep water bodies. Such water bodies rarely exist today on Earth, but were common through Earth geological history. The Dead Sea is currently the closest and probably the only modern analog for such environments. Recently, based on direct field measurements, laboratory experiments, direct numerical simulations, and sedimentological investigation, we have shown that there are fundamental differences between deposition at deep basins versus shallow basins, specifically in the seasonal to multi-annual scales and variations of halite solubility with depth. We have found that during the dry summer the epilimnion is warmer, saltier and undersaturated to halite, and that double diffusion flux delivers dissolved salt from the epilimnion into the hypolimnion, resulting in the continuously supersaturated hypolimnion and seasonally undersaturated epilimnion. Thus the stratified structure of the lake’s water column results in focusing of halite deposits into the deep parts of the basin and thinned deposits, or entirely dissolved, in the marginal parts. We further explore the role of laterally variable hydroclimatic conditions to the spatiotemporal dynamics of evaporitic deposits in a deep hypersaline waterbody. We focus on the role of diluted buoyant plume, overlaying part of the Dead Sea surface that laterally spreads from freshwater inflow. The lateral surface salinity variations results in lateral variations in evaporation, double diffusion fluxes, and hence evaporitic layer thickness. These can contribute to the study of the depositional environments of halite units throughout the geological record, following the concept of “the present as key to the past”. At the end of the talk, I will share some management ideas regarding the future of the Dead Sea.

Computing the structure of protein molecules was a grand challenge problem for half a century. The problem has now been solved using deep learning methods, with many calculated structures rivalling experiment in accuracy. In this talk I’ll describe how the field advanced to this point and the nature of the solution. I’ll also discuss some implications for other areas of structural biology, and broader implications for how science is done.

Neurons in the brain form complex networks of synaptic connections. These elaborate networks define the physical scaffold on which neural activity occurs, and shape the collective dynamics of groups of neurons. In this talk, I will present my work on understanding the structural design principles of neural networks, and the relations between their architecture and their functional properties. First, I will ask what are the structural features that shape the function of neural networks, and show we can learn these features from large ensembles of simulated networks. Second, I will discuss how a strong biological constraint on the structure of neural networks does not incur a computational cost, and may even be functionally beneficial. Third, I will show how we can build connectomes which capture both the structure and the function of real data, using a small number of simple biological features.
 

Zoom Link: https://weizmann.zoom.us/j/95406893197?pwd=REt5L1g3SmprMUhrK3dpUDJVeHlrZz09
Meeting ID: 954 0689 3197
Password: 750421