1 2 Vegetation is the leading emitter of volatile organic compounds (VOC), a key ingredient for ozone formation. The contribution of biogenic VOC (BVOC) emissions to regional ozone formation needs better quantification so that air quality regulators can effectively design emission control strategies. One of the key uncertainties for modeling BVOC emissions comes from the estimation of photosynthetically active radiation (PAR) reaching canopy. Satellite insolation retrieval data provide an alternative to prognostic meteorological models for representing the spatial and temporal variations of PAR. In this study, biogenic emission estimates generated with the MEGAN and BEIS biogenic emissions models using satellite or prognostic PAR are used to examine the contribution of BVOC to ozone in the United States. The Comprehensive Air Quality Model with Extensions (CAMx) is applied with Ozone Source Apportionment Technology (OSAT) and brute force zero-out sensitivity runs to quantify the biogenic contributions to ozone formation during May through September 2011. The satellite PAR retrievals are on average lower than modeled PAR and exhibit better agreement with SCAN and SURFRAD network measurements. Using satellite retrievals instead of modeled PAR reduces BEIS and MEGAN estimates of isoprene by an average of 3%-4% and 9%-12%, respectively. The simulations still overestimate observed ground-level isoprene concentrations by a factor of 1.1 for BEIS and 2.6 for MEGAN. The spatial pattern of biogenic ozone contribution diagnosed from OSAT differs from the brute force zero-out sensitivity results, with the former more smoothly distributed and the latter exhibiting peak impacts near metropolitan regions with intense anthropogenic NO x emissions. OSAT tends to apportion less ozone to biogenics as BVOC emissions increase, since that shifts marginal ozone formation toward more NO x-limited conditions. By contrast, zero-25 out source apportionment of ozone to biogenics increases with BVOC emissions. OSAT 26 simulations with BEIS show that BVOCs typically contribute 10-19% to regional ozone 27 concentrations at nonattainment receptor sites during episode days. 28
Predicting athlete injury risk has been a holy grail in sports medicine with little progress to date due to a variety of factors such as small sample sizes, significantly imbalanced data, and inadequate statistical approaches. Modeling approaches which are not able to account for the multiple interactions across factors can be misleading. We address the small sample size by collecting longitudinal data of NBA player injuries using publicly available data sources and develop a state of the art deep learning model, METIC, to predict future injuries based on past injuries, game activity, and player statistics. We evaluate model performance using metrics appropriate for imbalanced data and find that METIC performs significantly better than other traditional machine learning approaches. METIC uses feature learning to create interactive features which become meaningful in combination with each other. METIC can be used by practitioners and front offices to improve athlete management and reduce injury incidence, potentially saving sports teams millions in revenue due to reduced athlete injuries.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.