Princeton Public Schools

India experiences some of the highest fine particulate matter (PM2.5) concentrations globally. Understanding the spatiotemporal variations of PM2.5 and its source attribution requires robust air quality modeling supported by up-to-date emission inventories. Here we present the first WRF-Chem model evaluation and source attribution analysis for India for 2022, supported by updates in sectoral emission inventories and model parameterizations. We have incorporated an updated residential emission inventory reflecting recent transitions to cleaner fuels in Indian households and develop a plant-level inventory for Indian coal-fired power plants. Further major improvements include model updates to the secondary organic aerosol scheme and an improved representation of near-surface pollutant mixing. Collectively our improvements result in a simulation with annual PM2.5 bias of only 0.2±16.9 µg m−3 (0±31 %) across 288 surface monitoring sites in South Asia. We simulate an annual population-weighted (PW) mean PM2.5 concentration of 47.4 µg m−3. Compared to earlier studies, in 2022 India's residential sector remained the dominant source of PM2.5 in the Indo-Gangetic Plain, but ranked second nationally in PW mean PM2.5 concentrations (15 %, 7.3 µg m−3). Industrial emissions emerged as the largest domestic contributor to national PW mean PM2.5 (18 %, 8.6 µg m−3), with urban hotspots including Delhi and Mumbai. Power sector contributions ranked third nationally (13 %, 6.1 µg m−3) and were particularly influential in central India. Transboundary transport contributed more than any individual domestic source nationally (27 %, 12.8 µg m−3) with largest impacts in western India. These findings highlight the benefits of India's partial residential sector transition toward cleaner fuels, while underscoring the future benefits of controlling industrial and power sector air pollutant emissions.

Wildfire prediction plays an important role in ecosystem restoration, hazards prevention, and environmental biodiversity enhancement. Accurate predictions rely on the uncovering of the drivers behind fire events, such as interactions of climate, vegetation, and land cover variables. In this work we adopted a data driven machine learning approach to uncover the impact on wildfire probability prediction of different risk drivers. We downloaded MODIX 6.1 daily active fire data from NASA FIRMs, daily weather data and vegetation coverage data from re-analysis product of ERA-5 Land, and gridded daily Normalized Difference Vegetation Index (NDVI) from the NOAA (National Centers for Environmental Information). In this research, we applied two tree-based classification algorithms to predict the fire danger for the region centered around the border between British Columbia, Canada and Oregon, US, an area that has seen several large scale wildfires in year 2018 to 2021. Random Forest and Extreme Gradient Boosting are two main tree-based supervised algorithms to solve classification problems. We compared the accuracy and F1-score for the Random Forest model v.s. the XGBoost model. XGBoost slightly outperforms Random Forest in terms of F1-score and accuracy measure on test data. Further, we used ROC (receiver operating characteristic) curve to compare the false positive rate with the true positive rate for a model performance. XGBoost also outperforms Random Forest in terms of ROC-AUC measure. Besides the climate, the vegetation cover data, and the active fire data, we calculated three new variables that can be derived from the raw climate data and the satellite based NDVI index, i.e., total precipitation in the previous month, average temperature in the previous month, and the closest distance to the developed land. We compute the feature importance scores and the SHAP values to identify the marginal contribution of each feature. Sensitivity anal-ysis was conducted to study how the model performance of Random Forest or XGBoost is affected by the most important feature. Major fire events are commonly driven by either low relative humidity or high wind speed. Our results confirmed the significant marginal contributions from the surface pressure and the relative humidity. In addition, our model identified the proxy to human behavior, the distance to the developed land as the risk driver with potentially the most impact on wildfire predictions.

This comprehensive survey critically evaluates the integration of Physics-Informed Machine Learning (PIML), Neuro-Symbolic AI (NeSy), Temporal Foundation Models (TSFMs), and Reinforcement Learning (RL) for Physical Time Series (PTS) analysis. We synthesize methodological innovations across these four pillars, focusing on how their synergistic architectures can address the core challenges of nonlinearity, non-stationarity, and the need for physical plausibility inherent in real-world systems. The central thesis is that the path forward lies in hybrid models that combine the perceptual power of TSFMs, the physical consistency of PIML, the logical reasoning of NeSy, and the decision-making capabilities of RL. We review the state-of-the-art in bipartite and tripartite hybrid models and present a conceptual quadripartite framework for a truly integrated, intelligent system. To ground our analysis, this survey uses autonomous driving as a running case study, providing a concrete illustration of how these paradigms are integrated in practice. Finally, we identify the principal open research challenges in scalability, theoretical guarantees, and automated design, providing a roadmap for future research in this convergent field.

Life in a multicultural nation can be fraught. The United States is a case in point, with hostile tension between members of competing identity groups playing out today on streets, in offices, and across the media. Modern Americans assume that bridging race, gender, and class inequity is the stuff of public-constitutional-law. This assumption follows the lead of modern American lawyers, who migrated to this body of law just as historians, sociologists, and economists began to insist that the private law of tort was exclusively concerned with the accidental physical harms inevitable in a modern economy. According to this econostory, tort has no role to play in addressing individual dignitary harms inevitable in a multicultural society. Joseph Ranney's book, The Burdens of All: A Social History of American Tort Law, is one of the most deeply researched and reasoned entries in this canon, and was the rightful centerpiece of Marquette University Law School's 2023 Conference, Tort Law: What Can We Learn from Where It Has Been? In this response, originally offered as a keynote talk at that conference, I celebrate Ranney's achievement. But I also challenge him and fellow econohistorians of tort to reckon with the lost origins of American personal injury law. It turns out that in the pre-Industrial, Founding era, localized juries drawn from the community were often asked to intervene when women, enslaved people, and the poor claimed that neighbors degraded them because of their social characteristics. These claims were not uniformly successful, but they were a legal invitation for the community to examine its social norms and adjust them to stigmatize disfavored interpersonal conduct. As I recount, this private dignitary forum was dismantled when tort was reimagined as a quasiregulatory system to optimize resource allocation. The quest for individual * 2024-2025 Fellow, Princeton University Program in Law and Normative Thinking; Professor and Claire Ferguson-Carlson Fellow in Law, University of Iowa College of Law. Thank you to Joseph Kearney for inviting me to respond to Jay Ranney's wonderful book, to Jay and to Alex Lemann for helpful comments on an early draft, and to participants at the Michigan State University College of Law Faculty Workshop for generous feedback. Thanks also to Benjamin Marchant for outstanding research assistance, and to the editors at the Marquette Law Review for their careful and thoughtful assistance bringing these ideas to print. V23_TILLEY_POWER OF ALL (DO NOT DELETE) 1/9/2025 11:55 AM 102 MARQUETTE LAW REVIEW [108:ppp dignity was ultimately reassigned in the twentieth century to Article III courts, which updated and expanded their reading of the Constitution to integrate members of suspect classes more fully into public life. But while constitutional pronouncements can force system-level equity rules for schools, workplaces, police forces, and the like, they cannot touch the private worldviews of the people found in those systems. The Burdens of All acknowledges that American tort has always fluctuated to manage the social tensions of the day. I suggest it may be time for a twenty-first century fluctuation. Excavating tort's original purpose as an instrument of interpersonal dignity provides a template for building out a robust private law of social justice in the modern era.

Wildlife use and trade support the livelihoods of millions of people worldwide but also threaten thousands of species. Legal instruments, when effectively designed and implemented, can help regulate trade and mitigate negative impacts. However, activities along supply chains are rarely categorically legal or illegal, with considerable uncertainties regarding legality in the wildlife trade. These uncertainties can compromise the success of efforts to ensure, or improve, sustainability, but are often overlooked. Here, we categorize legal uncertainties in wildlife trade into three dimensions: institutional, operational, and perceptual. We explore their implications for sustainable management and discuss potential interventions to address them, drawing on examples from wildlife management and other sectors. Resolving these uncertainties can reduce unsustainable and illegal trade, strengthen traceability and enforcement, and promote equitable benefit-sharing among actors. Our findings offer actionable insights for policymakers, practitioners, and researchers to improve the clarity and effectiveness of wildlife trade management, advancing both conservation and socio-economic objectives.