Search for: All records

Abstract The export of the North Atlantic Deep Water (NADW) from the subpolar North Atlantic is known to affect the variability in the lower limb of the Atlantic meridional overturning circulation (AMOC). However, the respective impact from the transport in the upper NADW (UNADW) and lower NADW (LNADW) layers, and from the various transport branches through the boundary and interior flows, on the subpolar overturning variability remains elusive. To address this, the spatiotemporal characteristics of the circulation of NADW throughout the eastern subpolar basins are examined, mainly based on the 2014–20 observations from the transatlantic Overturning in the Subpolar North Atlantic Program (OSNAP) array. It reveals that the time-mean transport within the overturning’s lower limb across the eastern subpolar gyre [−13.0 ± 0.5 Sv (1 Sv ≡ 10⁶m³s⁻¹)] mostly occurs in the LNADW layer (−9.4 Sv or 72% of the mean), while the lower limb variability is mainly concentrated in the UNADW layer (57% of the total variance). This analysis further demonstrates a dominant role in the lower limb variability by coherent intraseasonal changes across the region that result from a basinwide barotropic response to changing wind fields. By comparison, there is just a weak seasonal cycle in the flows along the western boundary of the basins, in response to the surface buoyancy-induced water mass transformation. 

Recent reviews of science, technology, engineering, and mathematics (STEM) education conclude that engagement of undergraduate students in research generally broadens future participation in research and increased retention in STEM. Towards the goal of investing in a sustained and diverse atmospheric science research community, the Center for Climate and Aerosol Research (CCAR) at Portland State University (PSU) introduced a Research Experience for Undergraduates (REU) program in 2014 with the objective of providing atmospheric science summer research experiences to promising students in STEM disciplines from rural Northwest and Native American communities who would be unlikely to be otherwise exposed to such opportunities at their home institution. The PSU CCAR REU site is focused on student research in areas of atmospheric chemistry, physics, air quality, meteorology and climate change. For 10 weeks, students conduct research with an expert faculty mentor and participate in activities such as a short courses, faculty research seminars, and hands-on group workshops; academic professional and career development workshops; journal club activities; and opportunities for travel for student presentations at scientific conferences; and social activities. The program ends with a paper based on their summer research, which is presented via poster and oral presentations during our concluding CCAR symposium. Evaluation data from seven cohorts (2014-2021) of the CCAR REU (N = 70) was used to explore how science identity had changed over the course of the program, as well as what predicted positive increases in science identity. Change was assessed using paired-sample t-tests. To explore the predictors of change, we ran an exploratory stepwise regression where the difference score in science identity items from pre- to post-program was predicted by similar changes in knowledge, intrinsic motivation, extrinsic motivation, and career aspirations, demographic characteristics (e.g., age, gender), and mid-program satisfaction and met expectations. In this presentation, we present these findings along with supportive qualitative analyses and discuss their implications for undergraduate research programs in geoscience fields. 

We develop a rigorous mathematical analysis of zero-shot learning with attributes. In this setting, the goal is to label novel classes with no training data, only detectors for attributes and a description of how those attributes are correlated with the target classes, called the class-attribute matrix. We develop the first non-trivial lower bound on the worst-case error of the best map from attributes to classes for this setting, even with perfect attribute detectors. The lower bound characterizes the theoretical intrinsic difficulty of the zero-shot problem based on the available information---the class-attribute matrix---and the bound is practically computable from it. Our lower bound is tight, as we show that we can always find a randomized map from attributes to classes whose expected error is upper bounded by the value of the lower bound. We show that our analysis can be predictive of how standard zero-shot methods behave in practice, including which classes will likely be confused with others. 

Semi-supervised and unsupervised machine learning methods often rely on graphs to model data, prompting research on how theoretical properties of operators on graphs are leveraged in learning problems. While most of the existing literature focuses on undirected graphs, directed graphs are very important in practice, giving models for physical, biological or transportation networks, among many other applications. In this paper, we propose a new framework for rigorously studying continuum limits of learning algorithms on directed graphs. We use the new framework to study the PageRank algorithm and show how it can be interpreted as a numerical scheme on a directed graph involving a type of normalised graph Laplacian . We show that the corresponding continuum limit problem, which is taken as the number of webpages grows to infinity, is a second-order, possibly degenerate, elliptic equation that contains reaction, diffusion and advection terms. We prove that the numerical scheme is consistent and stable and compute explicit rates of convergence of the discrete solution to the solution of the continuum limit partial differential equation. We give applications to proving stability and asymptotic regularity of the PageRank vector. Finally, we illustrate our results with numerical experiments and explore an application to data depth. 

The Mixed-Reality Integrated Learning Environment (MILE) developed at Florida State University is a virtual reality based, inclusive and immersive e-learning environment that promotes engaging and effective learning interactions for a diversified learner population. MILE uses a large number of interactive Non-Player Characters (NPCs) to represent diverse research-based learner archetypes and groups, and to prompt and provide feedback for in situ teaching practice. The NPC scripts in MILE are written in Linden Scripting Language (LSL), and can be quite complex, creating a significant challenge in the development and maintenance of the system. To address this challenge, we develop NPC_GEN, an automatic NPC script generation tool that takes high-level NPC descriptions as input and automatically produces LSL scripts for NPCs. In this work, we introduce NPCDL, a language that we design for NPC_GEN to give high-level descriptions of NPCs, describe how NPC_GEN translates an NPCDL description into an LSL script, and report a user study of NPC_GEN. The results of our user study indicate that with minimal training, non-technical people are able to write and modify NPCDL descriptions, which can then be used to generate LSL scripts for the NPCs: the development and maintenance of NPCs is greatly simplified with NPC_GEN.