Search

The increase in the concentration of geological gas emissions in the atmosphere and particularly the increase of methane is considered by the majority of the scientific community as the main cause of global climate change. The main reasons that place methane at the center of interest, lie in its high global warming potential (GWP) and its lifetime in the atmosphere. Anthropogenic processes, like engineering geology ones, highly affect the daily profile of gasses in the atmosphere. Should direct measures be taken to reduce emissions of methane, immediate global warming mitigation could be achieved. Due to its significance, methane has been monitored by many space missions over the years and as of 2017 by the Sentinel-5P mission. Considering the above, we conclude that monitoring and predicting future methane concentration based on past data is of vital importance for the course of climate change over the next decades. To that end, we introduce a method exploiting state-of-the-art recurrent neural networks (RNNs), which have been proven particularly effective in regression problems, such as time-series forecasting. Aligned with the green artificial intelligence (AI) initiative, the paper at hand investigates the ability of different RNN architectures to predict future methane concentration in the most active regions of Texas, Pennsylvania and West Virginia, by using Sentinel-5P methane data and focusing on computational and complexity efficiency. We conduct several empirical studies and utilize the obtained results to conclude the most effective architecture for the specific use case, establishing a competitive prediction performance that reaches up to a 0.7578 mean squared error on the evaluation set. Yet, taking into consideration the overall efficiency of the investigated models, we conclude that the exploitation of RNN architectures with less number of layers and a restricted number of units, i.e., one recurrent layer with 8 neurons, is able to better compensate for competitive prediction performance, meanwhile sustaining lower computational complexity and execution time. Finally, we compare RNN models against deep neural networks along with the well-established support vector regression, clearly highlighting the supremacy of the recurrent ones, as well as discuss future extensions of the introduced work. [ABSTRACT FROM AUTHOR]

The Pell Grant is the foundational need-based student aid program in the United States, providing students of lower socio-economic status a pathway to afford college costs and educational expenses. Currently, over one-third of all U.S. undergraduate students receive Pell. This paper examines federal Pell assistance and institutional costs for students at the 38 publicly controlled regional Historically Black Colleges and Universities (HBCUs), which serve high average percentages of low-income students and students of color. By deploying the University of Alabama Education Policy Center's new Mission-Driven Classification System to enrollment, tuition and fees, and other costs metrics along with federal Pell and student loan data from the Integrated Postsecondary Education Data System (IPEDS), a more direct apples-to-apples comparison of the 38 public regional HBCUs to the 182 public regional non-HBCU in the same 19 southern states, is revealed, as are comparisons to the universe of 461 public regional universities nationally. This paper finds that America's most financially disadvantaged students rely on Pell Grants to alleviate financial constraints at public regional HBCUs, where 55 percent of students are Pell recipients, a rate 24 percent higher than their non-HBCU counterparts. Moreover, the data underscore an opportunity for Congress to construct a meaningful federal role in higher education by providing stable and sustainable funding for the Pell Grant program.

Background: Police violence is increasingly recognized as an urgent public health problem. Basic questions about police violence, however, remain unanswered, including which types of law enforcement agency are responsible for fatal police violence deaths. Methods: We estimated the proportion of police violence deaths in the U.S. (2013–2022) that were attributable to local, county, state, federal, or tribal police agencies, using mapping police violence data. We examined proportions overall, by decedent race/ethnicity, and by state. Results: Nationally, 60% of decedents were killed by municipal, 29% by county, 8% by state, and 3% by federal, police, with < 1% killed by tribal or other officers. These proportions varied by race/ethnicity, with 56% of Native American decedents killed by municipal police compared to 70–75% among other racially minoritized people. While municipal police were responsible for most deaths in most states, in the Southeast, county police predominated. In some Northeastern states (and Alaska), state police were responsible for > 40% of deaths. Conclusions: We identify wide geographic & racial/ethnic variation in the agencies responsible for fatal police violence. Findings suggest that the budgetary and infrastructural shifts required to prevent fatal police violence need to occur at multiple levels of government. [ABSTRACT FROM AUTHOR]

The Monongahela River Basin has an extensive history of fossil fuel development, including coal mining and natural gas extraction. In late summer 2008, total dissolved solids (TDS) concentrations exceeding the United States Environmental Protection Agency′s (EPA) secondary drinking water standards were detected. After determining the source, a voluntary discharge management plan (VDMP) was developed by the West Virginia Water Research Institute (WVWRI) and implemented by the coal industry (2010). Additional remediation actions included Pennsylvania's prohibition of produced wastewater in publicly owned treatment facilities (2011) and construction of a reverse osmosis treatment facility (2013). We used a locally weighted polynomial regression in conjunction with a segmented regression to assess the discharge and concentration trends/changepoints for bromide, chloride, sulfate, and total dissolved solids at various locations relative to the three remedial actions. We detected significant (α < 0.05) positive trends for discharge and significant negative trends for bromide, chloride, sulfate, and total dissolved solids. In conjunction, we also detected 1–4 changepoints within each model. Additionally, a linear mixed effects model containing discharge and remedial actions was used to measure the effectiveness of each remediation action in reducing TDS over time. Of the three remedial actions, the VDMP by itself was effective in maintaining river sulfate and TDS levels below the secondary drinking water standards (−0.12, p-value = 0.002). The combination of the VDMP with Pennsylvania's produced water prohibition (−0.16, p-value < 0.001) and the combination of the VDMP with the reverse osmosis treatment facility (−0.19, p-value < 0.001) were also effective. The use of all three remedial actions produced the strongest effect (−0.37, p-value < 0.001) Since the implementation of these changes, primarily the VDMP which encompasses most of the watershed, TDS in the Monongahela has not exceeded the EPA′s secondary drinking water standards. Future management decisions should include efforts to further expand the VDMP and to monitor changes in land use or severe changes in discharge. [ABSTRACT FROM AUTHOR]

The Middle Devonian Marcellus Formation of North America is the most prolific hydrocarbon play in the Appalachian basin, the second largest producer of natural gas in the United States, and one of the most productive gas fields in the world. Regional differences in Marcellus fluid chemistry reflect variations in thermal maturity, migration, and hydrocarbon alteration. These differences define specific wet gas/condensate and dry gas production in the basin. Marcellus gases co-produced with condensate in southwest Pennsylvania and northwest West Virginia are mixtures of residual primary-associated gases generated in the late oil window and postmature secondary hydrocarbons generated from oil cracking in the wet gas window. Correlation of API gravity and C7 expulsion temperatures, high heptane and isoheptane ratios, and the gas geochemical data confirm that the Marcellus condensates formed through oil cracking. Respective low toluene/nC7 and high nC7/methylcyclohexane ratios indicate selective depletion of low-boiling point aromatics and cyclic light saturates in all samples, suggesting that water washing and gas stripping altered the fluids. These alterations may be related to deep migration of hot basinal brines. Dry Marcellus gases produced in northeast Pennsylvania and northcentral West Virginia are mixtures of overmature methane largely cracked from refractory kerogen and ethane and propane cracked from light oil and wet gas. Carbon and hydrogen isotope distributions are interpreted to indicate (1) mixing of hydrocarbons of different thermal maturities, (2) high temperature Rayleigh fractionation of wet gas during redox reactions with transition metals and formation water, (3) isotope exchange between methane and water, and, possibly, (4) thermodynamic equilibrium conditions within the reservoirs. Evidence for thermodynamic equilibrium in the dry gases includes measured molecular proportions (C1/(C1 − C5) = 0.96 to 0.985) and δ13C1 values significantly greater than δ13CKEROGEN. Noble gas systematics support the interpretation of hydrocarbon–formation water interactions, constrain the high thermal maturity of the hydrocarbon fluids, and provide a method of quantifying gas retention versus expulsion in the reservoirs. [ABSTRACT FROM AUTHOR]

This article presents a method for cave surveying in total darkness using an autonomous aerial vehicle equipped with a depth camera for mapping, downward-facing camera for state estimation, and forward and downward lights. Traditional methods of cave surveying are labor-intensive and dangerous due to the risk of injury when operating in darkness, and the potential structural instability of the subterranean environment. Although these dangers can be mitigated by deploying robots to map dangerous passages and voids, real-time feedback is often needed to operate robots safely and efficiently. Few state-of-the-art, high-resolution perceptual modeling techniques attempt to reduce their high bandwidth requirements to work well with low bandwidth communication channels. To bridge this gap in the state-of- the-art, this work compactly represents sensor observations as Gaussian mixture models and maintains a local occupancy grid map for a motion planner that greedily maximizes an information-theoretic objective function. The approach accommodates both limited field of view (FoV) depth cameras and larger FoV LiDAR sensors and is extensively evaluated in long duration simulations on an embedded PC. An aerial system is leveraged to demonstrate the repeatability of the approach in a flight arena as well as the effects of communication dropouts. Finally, the system is deployed in Laurel Caverns, a commercially owned and operated cave in southwestern Pennsylvania, USA, and a wild cave in West Virginia, USA. Videos of the simulation and hardware results are available at https://youtu.be/iwi3p7IENjE and https://youtu.be/H8MdtJ5VhyU. [ABSTRACT FROM AUTHOR]

We use the synthetic control method to determine the economic impact of shale exploitation on Ohio, Pennsylvania, and West Virginia. Estimation results are mixed. The shale development decreased the poverty rate and increased the employment growth rate in Pennsylvania and West Virginia in the short run. Top oil and gas producing counties in West Virginia also experienced short‐term personal income growth due to fracking. However, most of the positive impacts disappeared a few years after the initial boom periods. The shale development did not bring significant economic benefits to Ohio. Further, shale drilling activities exert a potential long‐term negative effect on population growth in all three states. [ABSTRACT FROM AUTHOR]