Universidad Santa Lucía

Abstract Recently, a number of state-of-the-art surrogate machine learning (ML) models have been designed for global weather and climate prediction, which have been trained using reanalysis data products. Reanalysis data products are constructed using numerical model simulations that combine numerical integration of partial differential equations and parameterization schemes. These products are typically only archived and made available using coarsened spatial and temporal resolutions. This study explores the impact of the numerical generation methods used to produce the training datasets and the temporal resolution of those datasets on machine learning surrogate models. Using the nonlinear vector autoregression (NVAR) machine as an explainable ML technique, simple dynamical systems are emulated with ML models trained on data produced by three classical numerical integration schemes. NVAR is validated as a skillful ML method, capable of producing accurate predictions and, more importantly, reconstructing both the underlying dynamics and the numerical integration scheme used to generate the training data. However, the machine fails to generalize predictions on unseen test data generated by different numerical integration schemes, despite the underlying dynamical system being the same. This result provides a word of caution for the growing field of machine learning emulation of weather and climate dynamics. Furthermore, we illustrate using NVAR that training on temporally coarsened data may increase the required complexity of ML models and potentially introduce new numerical challenges. Finally, we discover that empirical integration schemes with arbitrary time-stepping sizes can be constructed directly from the data, which implies a potential for the development of empirical numerical integration schemes.

At the Electron Stretcher Facility ELSA of Bonn University, an upgrade of the maximum stored beam current from 20 mA to 200 mA is planned. The storage ring operates applying a fast energy ramp of 6 GeV/s from 1.2 GeV to 3.5 GeV and afterwards a slow beam extraction over a few seconds to the hadron physics experiments. The intended upgrade is mainly limited by the excitation of multibunch instabilities. As a countermeasure, we successfully commissioned state-of-the-art bunch by bunch feedback systems in the longitudinal and the two transverse dimensions. To achieve the maximum feedback gain, the beam phase with respect to the reference signal and the synchrotron frequency has to stay constant. A high performance low level RF system has been commissioned which is able to stabilize voltage and phase of the RF station while in addition taking care of resonator resonance tuning using a fast FPGA based fully digital approach.

Hybridization is common in both animals and plants and can lead to a diverse array of outcomes ranging from the generation of new ecotypes or species to the breakdown of morphological differences. Here, we explore the extent of hybridization in the three currently recognized New World Rhizophora species-R. mangle, R. racemosa, and the putative hybrid species R. harrisonii. • We assayed variation across the three recognized Rhizophora species using two noncoding chloroplast (cpDNA), two flanking microsatellite regions (FMRs), and six microsatellite loci. • Gene genealogies of cpDNA and FMRs showed a strong phylogeographic break across the Central American Isthmus, but little relationship to recognized species boundaries. Instead, individuals collected in the same ocean basin and classified as R. mangle and R. racemosa by morphological characteristics were more closely related to each other than with similar looking individuals collected in the other ocean basin. Nonetheless, there were low, yet significant differences at microsatellite loci among co-occurring populations of R. mangle and R. racemosa in both ocean basins, suggesting that two taxonomic groups coexist. However, we found no genetic evidence that R. harrisonii was a hybrid species. Rather, R. harrisonii appears to represent a morphotype produced by ongoing hybridization and backcrossing between R. mangle and R. racemosa. • Our data support ancient and persistent introgressive hybridization among new world Rhizophora and argue for a full revision of the systematic relationships of the group based on much finer morphological, ecological, and genetic analyses.

Our topic became: pacing the curriculum in the context of school realities. Our 'study' stemmed from our long conversations about teaching and our observations about the changes we had each made over the years to adjust to the ebb and flow of the school year. (We believe that this is "action research" because it has been all about "making changes and observing their effects"-- but it may be one of the longest running action-research studies in that it has taken us over thirty years to articulate our conclusions.) From those conversations we developed a prototype of factors, conditions, variables, that we personally found tend to influence pacing the curriculum in the day, week, month, term, year. We found that the changes we had both made with regard to pacing the curriculum were almost identical. For a "reality check" on our findings, we then surveyed twenty-four other secondary teachers, teachers with reputations for being strong teachers to see if there was agreement about the importance of the factors we had identified and to see what else they might add to this ongoing conversation. Based on our reflection about our own experiences, and our survey of twenty-four of our peers, we found general support for the idea that the factors we identified have strong external influences on the pacing and the rhythm of the curriculum. While we think there is value for teachers in considering the factors that we identified, we felt that the greater richness of our study came in the written comments by the teachers we surveyed. On the basis of their remarks we found evidence that effective teachers have adapted to these factors to mitigate against their becoming negative factors, and that at least certain teachers have a rich repository of knowledge about the art of pacing the curriculum.

Brain oedema is a major cause of early death after stroke. A 10% solution of glycerol is a hyperosmolar agent that is claimed to reduce brain oedema. To determine whether I.V. glycerol treatment in acute stroke, either ischaemic or haemorrhagic, influences death rates and functional outcome in the short or long term and whether the treatment is safe. The Cochrane Stroke Group Trials Register was searched, conference proceedings were screened and some trialists were personally contacted. All completed, randomised and quasi-randomized, controlled, published and unpublished comparisons, evaluating clinical outcome in which intravenous (I.V.) glycerol treatment was initiated within the first days after stroke onset. Two reviewers independently applied the inclusion criteria, assessed the trial quality and extracted data and this was checked with all co-reviewers. Death from all causes, functional outcome and adverse effects were analysed. Eleven completed, randomised trials comparing I.V. glycerol and control were considered. Analysis of death during the scheduled treatment period for acute ischaemic and/or haemorrhagic stroke was possible in ten trials where 482 glycerol treated patients were compared with 463 control patients. Glycerol was associated with a non-significant reduction in the odds of death within the scheduled treatment period (OR 0.78, 95% Confidence Intervals 0.58 - 1.06). Among patients with definite or probable ischaemic stroke, glycerol was associated with a significant reduction in the odds of death during the scheduled treatment period (odds ratio 0.65, 95% CI 0.44-0.97). However, at the end of the scheduled follow up period, there was no significant difference in the odds of death (odds ratio 0.98, 95% CI 0.73-1.31). Functional outcome was reported in only two studies but there were non-significantly more patients who had a good outcome at the end of scheduled follow up (odds ratio 0.73, 95% CI 0.37-1.42). Haemolysis seems to be the only relevant adverse effect of glycerol treatment. This systematic review suggests a favourable effect of glycerol treatment on short term survival in patients with probable or definite ischaemic stroke but the magnitude of the treatment effect may be minimal (as low as a 3% reduction in odds). Due to the relatively small number of patients and that the trials have been performed in the pre-CT era, the results must be interpreted cautiously. The lack of evidence of benefit in long term survival does not support the routine or selective use of glycerol treatment in patients with acute stroke.