Search for: All records

Crystals retain an imprint of the dynamic changes within a magma reservoir and hence contain invaluable information about the evolving conditions inside volcanic plumbing systems. However, instead of telling a single, simple story, they comprise overprinted evidence of numerous processes relating to temperature, pressure and composition that drive crystal precipitation and dissolution in magmatic systems. To decipher these different elements in the story that crystals tell, we attempt to identify the observational signatures of a simple, yet ubiquitous process: crystal precipitation and dissolution during magma cooling. To isolate this process in a complex magmatic system with intricate dynamic feedbacks, we assume that synthetic crystals precipitate and dissolve rapidly in response to deviations from thermodynamic equilibrium. In our crystalline‐scale simulations, synthetic crystals drag along the cooler‐than‐ambient melt in which they precipitated and can drive a temperature‐dependent, crystal‐driven convection. We analyze the non‐dimensional conditions for this coupled convection and record the heterogeneous thermal histories that synthetic crystals in this flow regime experience. We show that many synthetic crystals dissolve, loosing their thermal record of the convection. Based on our findings, we suggest that heterogeneity in the thermal history of crystals is more indicative of local, crystal‐scale processes than the overall, system‐wide cooling trend.

 

Derivation of partial charges in small and large scale molecular systems is important for modeling of various experimental and theoretical properties like dipole moments, auto-correlation functions, charge disparity, understanding of dispersion, benchmark of classical MD simulations and electrostatic potential energy surface mapping. A correspondence between theoretical calculations (based on single/small number of molecules) is usually established with macroscopic IR/Raman spectra or dipole moment measurements. Such comparisons are indirect and lack a fine mapping of electrostatic potential from theory to experiment. In a new approach developed as the experimental part of this work, partial charges are calculated from crystallographic model refinement. The experimental method exhibits a satisfactory correspondence with partial charges obtained using quantum chemistry calculations. Particularly, gas phase partial charges from CHELPG method and condensed phase Lowdin charges correlate well and validate this experimental method. 

This paper investigates online stochastic planning for problems with large factored state and action spaces. One promising approach in recent work estimates the quality of applicable actions in the current state through aggregate simulation from the states they reach. This leads to significant speedup, compared to search over concrete states and actions, and suffices to guide decision making in cases where the performance of a random policy is informative of the quality of a state. The paper makes two significant improvements to this approach. The first, taking inspiration from lifted belief propagation, exploits the structure of the problem to derive a more compact computation graph for aggregate simulation. The second improvement replaces the random policy embedded in the computation graph with symbolic variables that are optimized simultaneously with the search for high quality actions. This expands the scope of the approach to problems that require deep search and where information is lost quickly with random steps. An empirical evaluation shows that these ideas significantly improve performance, leading to state of the art performance on hard planning problems.