Search

Constraining planet formation based on the atmospheric composition of exoplanets is a fundamental goal of the exoplanet community. Existing studies commonly try to constrain atmospheric abundances, or to analyze what abundance patterns a given description of planet formation predicts. However, there is also a pressing need to develop methodologies that investigate how to transform atmospheric compositions into planetary formation inferences. In this study we summarize the complexities and uncertainties of state-of-the-art planet formation models and how they influence planetary atmospheric compositions. We introduce a methodology that explores the effect of different formation model assumptions when interpreting atmospheric compositions. We apply this framework to the directly imaged planet HR 8799e. Based on its atmospheric composition, this planet may have migrated significantly during its formation. We show that including the chemical evolution of the protoplanetary disk leads to a reduced need for migration. Moreover, we find that pebble accretion can reproduce the planet's composition, but some of our tested setups lead to too low atmospheric metallicities, even when considering that evaporating pebbles may enrich the disk gas. We conclude that the definitive inversion from atmospheric abundances to planet formation for a given planet may be challenging, but a qualitative understanding of the effects of different formation models is possible, opening up pathways for new investigations., Accepted for publication in ApJ, base formation inversion code is available at: https://gitlab.com/mauricemolli/formation-inversion

Drawing on comparative resource curse literature and American literature on the determinants of corruption, we argue that the impact of natural resource extraction on corruption outcomes is state-dependent. That is, in environments where corruption is already high, natural resource windfalls allow political actors and economic elites to take advantage of state brokerage, further increasing corruption. However, in previously less-corrupt states, increased natural resource extraction will not induce corruption. We rely on hierarchical linear models to interpret federal corruption convictions data for the fifty American states between 1976 and 2012 and employ generalized method of moments estimators to account for potential endogeneity. The findings are robust to alternative specifications and have implications for the management of new resource extraction opportunities.

We investigate how radiatively driven heating and cooling in the upper atmosphere (at pressures below 1 bar) influences the interior temperature profile (at pressures between 1 to 700 bar) by means of dynamical heat transport. To achieve this goal, we perform fully coupled 3D-radiation-hydrodymamical models with the new full RT 3D climate model MITgcm/ExoRadPRT for WASP-43 b and HD209458 b. We show in our simulations under which conditions the interior temperature profile converges to a hot deep adiabat. Furthermore, we show if differences occur between the non inflated WASP-43 b and the inflated HD209458 b due to different flow structures at depth for similar irradiation.

Disfluency detection models now approach high accuracy on English text. However, little exploration has been done in improving the size and inference time of the model. At the same time, automatic speech recognition (ASR) models are moving from server-side inference to local, on-device inference. Supporting models in the transcription pipeline (like disfluency detection) must follow suit. In this work we concentrate on the disfluency detection task, focusing on small, fast, on-device models based on the BERT architecture. We demonstrate it is possible to train disfluency detection models as small as 1.3 MiB, while retaining high performance. We build on previous work that showed the benefit of data augmentation approaches such as self-training. Then, we evaluate the effect of domain mismatch between conversational and written text on model performance. We find that domain adaptation and data augmentation strategies have a more pronounced effect on these smaller models, as compared to conventional BERT models., INTERSPEECH 2021

Mobility restrictions during the COVID-19 pandemic ostensibly prevented the public from transmitting the disease in public places, but they also hampered outdoor recreation, despite the importance of blue-green spaces (e.g., parks and natural areas) for physical and mental health. We assess whether restrictions on human movement, particularly in blue-green spaces, affected the transmission of COVID-19. Our assessment uses a spatially resolved dataset of COVID-19 case numbers for 848 administrative units across 153 countries during the first year of the pandemic (February 2020 to February 2021). We measure mobility in blue-green spaces with planetary-scale aggregate and anonymized mobility flows derived from mobile phone tracking data. We then use machine learning forecast models and linear mixed-effects models to explore predictors of COVID-19 growth rates. After controlling for a number of environmental factors, we find no evidence that increased visits to blue-green space increase COVID-19 transmission. By contrast, increases in the total mobility and relaxation of other non-pharmaceutical interventions such as containment and closure policies predict greater transmission. Ultraviolet radiation stands out as the strongest environmental mitigant of COVID-19 spread, while temperature, humidity, wind speed, and ambient air pollution have little to no effect. Taken together, our analyses produce little evidence to support public health policies that restrict citizens from outdoor mobility in blue-green spaces, which corroborates experimental studies showing low risk of outdoor COVID-19 transmission. However, we acknowledge and discuss some of the challenges of big data approaches to ecological regression analyses such as this, and outline promising directions and opportunities for future research.

Better relaxing lockdown together Even during a pandemic, all countries—even islands—are dependent in one way or another on their neighbors. Without coordinated relaxation of nonpharmaceutical interventions (NPIs) among the most closely connected countries, it is difficult to envisage maintaining control of infectious viruses such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Ruktanonchai et al. used mobility data from smartphones to estimate movements between administrative units across Europe before and after the implementation of NPIs for coronavirus disease 2019 (COVID-19). Modeling disease dynamics under alternative scenarios of countries releasing NPIs, in particular stay-at-home orders, showed that if countries do not coordinate their NPIs when they relax lockdown, resurgence of disease occurs sooner. Coordination of on-off NPIs would significantly increase their effectiveness at reducing transmission across Europe. Science , this issue p. 1465

Introduction Alzheimer’s Disease (AD) is the most common form of dementia (1) and the 6th leading cause of death in the U.S. (2). Additionally, it is estimated that currently 5 million individuals are living with AD in the U.S. and this number is expected to triple by 2050 (3). Decades of research has identified various risk factors for AD and a wide array of cardiometabolic risk factors, such as hypertension, type 2 diabetes, metabolic syndrome and obesity (4, 5, 6, 7). Several behavioral factors have also been suggested to play a protective role against AD, including Mediterranean-type diet (8), physical activity (9), and cognitively stimulating activities (10, 11). More recently, growing evidence has pointed sleep deficiency as a modifiable risk factor for AD. (12). Methods The purpose of our study was to determine if sleep and napping were associated with AD mortality. We used data from the NIH-AARP Diet and Health Study. Sleep duration and napping were self-reported and AD death were ascertained via linkage to the National Death Index. Of 566,398 members (aged 50-71) who completed baseline questionnaire in 1995-1996, 337,373 participants were included (195,656 men and 137,018 women). Results Long sleep and napping were both associated with increased AD mortality. Specifically, 9+ hours of sleep was associated with 50% increase (Hazard Ratio 1.50, 95% CI (1.17-1.92)) in AD mortality when compared 7-8 hours, while napping for 1+ hours was associated with 29% increase (1.29 (1.08, 1.55)) when compared to no napping. Results appeared to be stronger in men and remained after removing AD deaths within first 5 years after baseline. Conclusion The results of this study showed that 9+ hours sleep duration was associated with a 50% increase in AD mortality risk when compared to people who slept between 7 and 8 hours, particularly in men. Additional longitudinal studies that follow middle-aged adults through death and include objective measures of sleep and AD related biomarkers are needed to further elucidate the mechanisms contributing to cognitive decline and AD death. Such studies would be useful for informing interventions designed to reduce risk of AD and AD-related death. Support (If Any)

Changes to citizenship regimes help explain differences in tax structure in Brazil and India. Citizenship regimes change by mobilizing new collective identities, articulating substantive demands, and institutionalizing group linkages to public life. When excluded groups mobilize and gain access, they provide new sources of state legitimacy, allowing states to expand their capacity, for example in tax. Changes to tax can be evaluated in terms of levels of revenues, degrees of progressivity, and the universality of application of tax across sectors and regions. Since the 1970s in Brazil and India, excluded groups gained access to citizenship regimes and deepened democracy, but differences in collective identities, demands, and mechanisms of linkage played out in differences in state capacity.

With each new period of international insertion, the Brazilian state has sought capacity to promote dynamic sectors, evident in major, coincident changes to two institutions: federalism and tax. Coincident reforms mobilize new resources and achieve accommodations among elites leading different jurisdictions and levels of government. To make sense of these accommodations, this chapter explores the social relations of fiscal and federal affairs—fiscal federal sociology. In contrast to coincident reforms to advance state capacity, divergent reforms occur when fiscal agreements and/or accommodations among regional elites breakdown. The historical evolution of coincident and divergent reforms produces the current paradox of a Brazilian state that shows great capacity in some areas and some periods, but has experienced long periods of stagnation and regression.