Search for: All records

Shergottites are mafic to ultramafic igneous rocks that represent the majority of known Martian meteorites. They are subdivided into gabbroic, poikilitic, basaltic, and olivine–phyric categories based on differences in mineralogy and textures. Their geologic contexts are unknown, so analyses of crystal sizes and preferred orientations have commonly been used to infer where shergottites solidified. Such environments range from subsurface cumulates to shallow intrusives to extrusive lava flows, which all have contrasting implications for interactions with crustal material, cooling histories, and potential in situ exposure at the surface. In this study, we present a novel three‐dimensional (3‐D) approach to better understand the solidification environments of these samples and improve our knowledge of shergottites' geologic contexts. Shape preferred orientations of most phases and crystal size distributions of late‐forming minerals were measured in 3‐D using X‐ray computed tomography (CT) on eight shergottites representing the gabbroic, poikilitic, basaltic, and olivine–phyric categories. Our analyses show that highly anisotropic, rod‐like pyroxene crystals are strongly foliated in the gabbroic samples but have a weaker foliation and a mild lineation in the basaltic sample, indicating a directional flow component in the latter. Star volume distribution analyses revealed that most phases (maskelynite, pyroxene, olivine, and oxides/sulfides) preserve a foliated texture with variable strengths, and that the phases within individual samples are strongly to moderately aligned with respect to one another. In combination with relative cooling rates during the final stages of crystallization determined from interstitial oxide/sulfide crystal size distribution analyses, these results indicate that the olivine–phyric samples were emplaced as shallow intrusives (e.g., dikes/sills) and that the gabbroic, poikilitic, and basaltic samples were emplaced in deeper subsurface environments.

 

Data driven individualized decision making problems have received a lot of attentions in recent years. In particular, decision makers aim to determine the optimal Individualized Treatment Rule (ITR) so that the expected speci ed outcome averaging over heterogeneous patient-speci c characteristics is maximized. Many existing methods deal with binary or a moderate number of treatment arms and may not take potential treatment e ect structure into account. However, the e ectiveness of these methods may deteriorate when the number of treatment arms becomes large. In this article, we propose GRoup Outcome Weighted Learning (GROWL) to estimate the latent structure in the treatment space and the op- timal group-structured ITRs through a single optimization. In particular, for estimating group-structured ITRs, we utilize the Reinforced Angle based Multicategory Support Vec- tor Machines (RAMSVM) to learn group-based decision rules under the weighted angle based multi-class classi cation framework. Fisher consistency, the excess risk bound, and the convergence rate of the value function are established to provide a theoretical guaran- tee for GROWL. Extensive empirical results in simulation studies and real data analysis demonstrate that GROWL enjoys better performance than several other existing methods. 

Deep Learning for Time-series plays a key role in AI for healthcare. To predict the progress of infectious disease outbreaks and demonstrate clear population-level impact, more granular analyses are urgently needed that control for important and potentially confounding county-level socioeconomic and health factors. We forecast US county-level COVID-19 infections using the Temporal Fusion Transformer (TFT). We focus on heterogeneous time-series deep learning model prediction while interpreting the complex spatiotemporal features learned from the data. The significance of the work is grounded in a real-world COVID-19 infection prediction with highly non-stationary, finely granular, and heterogeneous data. 1) Our model can capture the detailed daily changes of temporal and spatial model behaviors and achieves better prediction performance compared to other time-series models. 2) We analyzed the attention patterns from TFT to interpret the temporal and spatial patterns learned by the model. 3) We collected around 2.5 years of socioeconomic and health features for 3142 US counties, such as observed cases, and a number of static (age distribution and health disparity) and dynamic features (vaccination, disease spread, transmissible cases, and social distancing). Using the proposed framework, we have shown that our model can learn complex interactions. Interpreting different impacts at the county level would be crucial for understanding the infection process that can help effective public health decision-making. 

Forest dynamics in arid and semiarid regions are sensitive to water availability, which is becoming increasingly scarce as global climate changes. The timing and magnitude of precipitation in the semiarid southwestern U.S. (“Southwest”) has changed since the 21st century began. The region is projected to become hotter and drier as the century proceeds, with implications for carbon storage, pest outbreaks, and wildfire resilience. Our goal was to quantify the importance of summer monsoon precipitation for forested ecosystems across this region. We developed an isotope mixing model in a Bayesian framework to characterize summer (monsoon) precipitation soil water recharge and water use by three foundation tree species (Populus tremuloides [aspen], Pinus edulis [piñon], and Juniperus osteosperma [Utah juniper]). In 2016, soil depths recharged by monsoon precipitation and tree reliance on monsoon moisture varied across the Southwest with clear differences between species. Monsoon precipitation recharged soil at piñon-juniper (PJ) and aspen sites to depths of at least 60 cm. All trees in the study relied primarily on intermediate to deep (10- 60 cm) moisture both before and after the onset of the monsoon. Though trees continued to primarily rely on intermediate to deep moisture after the monsoon, all species increased reliance on shallow soil moisture to varying degrees. Aspens increased reliance on shallow soil moisture by 13% to 20%. Utah junipers and co-dominant ñons increased their reliance on shallow soil moisture by about 6% to 12%. Nonetheless, approximately half of the post-monsoon moisture in sampled piñon (38-58%) and juniper (47- 53%) stems could be attributed to the monsoon. The monsoon contributed lower amounts to aspen stem water (24-45%) across the study area with the largest impacts at sites with recent precipitation. Therefore, monsoon precipitation is a key driver of growing season moisture that semiarid forests rely on across the Southwest. This monsoon reliance is of critical importance now more than ever as higher global temperatures lead to an increasingly unpredictable and weaker North American Monsoon.