Kaiguang Zhao

Increased stray gas abundance in a subset of drinking water wells near Marcellus shale gas extraction

Horizontal drilling and hydraulic fracturing are transforming energy production, but their potential environmental effects remain controversial. We analyzed 141 drinking water wells across the Appalachian Plateaus physiographic province of northeastern Pennsylvania, examining natural gas concentrations and isotopic signatures with proximity to shale gas wells. Methane was detected in 82% of drinking water samples, with average concentrations six times higher for homes <1 km from natural gas wells (P = 0.0006). Ethane was 23 times higher in homes <1 km from gas wells (P = 0.0013); propane was detected in 10 water wells, all within approximately 1 km distance (P = 0.01). Of three factors previously proposed to influence gas concentrations in shallow groundwater (distances to gas wells, valley bottoms, and the Appalachian Structural Front, a proxy for tectonic deformation), distance to gas wells was highly significant for methane concentrations (P = 0.007; multiple regression), whereas distances to valley bottoms and the Appalachian Structural Front were not significant (P = 0.27 and P = 0.11, respectively). Distance to gas wells was also the most significant factor for Pearson and Spearman correlation analyses (P < 0.01). For ethane concentrations, distance to gas wells was the only statistically significant factor (P < 0.005). Isotopic signatures (δ13C-CH4, δ13C-C2H6, and δ2H-CH4), hydrocarbon ratios (methane to ethane and propane), and the ratio of the noble gas 4He to CH4 in groundwater were characteristic of a thermally postmature Marcellus-like source in some cases. Overall, our data suggest that some homeowners living <1 km from gas wells have drinking water contaminated with stray gases.

Geochemical evidence for possible natural migration of Marcellus Formation brine to shallow aquifers in Pennsylvania

The debate surrounding the safety of shale gas development in the Appalachian Basin has generated increased awareness of drinking water quality in rural communities. Concerns include the potential for migration of stray gas, metal-rich formation brines, and hydraulic fracturing and/or flowback fluids to drinking water aquifers. A critical question common to these environmental risks is the hydraulic connectivity between the shale gas formations and the overlying shallow drinking water aquifers. We present geochemical evidence from northeastern Pennsylvania showing that pathways, unrelated to recent drilling activities, exist in some locations between deep underlying formations and shallow drinking water aquifers. Integration of chemical data (Br, Cl, Na, Ba, Sr, and Li) and isotopic ratios (87Sr/86Sr, 2H/H, 18O/16O, and 228Ra/226Ra) from this and previous studies in 426 shallow groundwater samples and 83 northern Appalachian brine samples suggest that mixing relationships between shallow ground water and a deep formation brine causes groundwater salinization in some locations. The strong geochemical fingerprint in the salinized (Cl > 20 mg/L) groundwater sampled from the Alluvium, Catskill, and Lock Haven aquifers suggests possible migration of Marcellus brine through naturally occurring pathways. The occurrences of saline water do not correlate with the location of shale-gas wells and are consistent with reported data before rapid shale-gas development in the region; however, the presence of these fluids suggests conductive pathways and specific geostructural and/or hydrodynamic regimes in northeastern Pennsylvania that are at increased risk for contamination of shallow drinking water resources, particularly by fugitive gases, because of natural hydraulic connections to deeper formations.

Ground Filtering Algorithms for Airborne LiDAR Data: A Review of Critical Issues

This paper reviews LiDAR ground filtering algorithms used in the process of creating Digital Elevation Models. We discuss critical issues for the development and application of LiDAR ground filtering algorithms, including filtering procedures for different feature types, and criteria for study site selection, accuracy assessment, and algorithm classification. This review highlights three feature types for which current ground filtering algorithms are suboptimal, and which can be improved upon in future studies: surfaces with rough terrain or discontinuous slope, dense forest areas that laser beams cannot penetrate, and regions with low vegetation that is often ignored by ground filters.

Mapping urban pipeline leaks: methane leaks across Boston.

Natural gas is the largest source of anthropogenic emissions of methane (CH(4)) in the United States. To assess pipeline emissions across a major city, we mapped CH(4) leaks across all 785 road miles in the city of Boston using a cavity-ring-down mobile CH(4) analyzer. We identified 3356 CH(4) leaks with concentrations exceeding up to 15 times the global background level. Separately, we measured δ(13)CH(4) isotopic signatures from a subset of these leaks. The δ(13)CH(4) signatures (mean = -42.8‰ ± 1.3‰ s.e.; n = 32) strongly indicate a fossil fuel source rather than a biogenic source for most of the leaks; natural gas sampled across the city had average δ(13)CH(4) values of -36.8‰ (± 0.7‰ s.e., n = 10), whereas CH(4) collected from landfill sites, wetlands, and sewer systems had δ(13)CH(4) signatures ~20‰ lighter (μ = -57.8‰, ± 1.6‰ s.e., n = 8). Repairing leaky natural gas distribution systems will reduce greenhouse gas emissions, increase consumer health and safety, and save money.

Natural Gas Pipeline Leaks Across Washington, DC

Pipeline safety in the United States has increased in recent decades, but incidents involving natural gas pipelines still cause an average of 17 fatalities and

Analysis of driver injury severity in rural single-vehicle crashes

133 M in property damage annually. Natural gas leaks are also the largest anthropogenic source of the greenhouse gas methane (CH4) in the U.S. To reduce pipeline leakage and increase consumer safety, we deployed a Picarro G2301 Cavity Ring-Down Spectrometer in a car, mapping 5893 natural gas leaks (2.5 to 88.6 ppm CH4) across 1500 road miles of Washington, DC. The δ(13)C-isotopic signatures of the methane (-38.2‰ ± 3.9‰ s.d.) and ethane (-36.5 ± 1.1 s.d.) and the CH4:C2H6 ratios (25.5 ± 8.9 s.d.) closely matched the pipeline gas (-39.0‰ and -36.2‰ for methane and ethane; 19.0 for CH4/C2H6). Emissions from four street leaks ranged from 9200 to 38,200 L CH4 day(-1) each, comparable to natural gas used by 1.7 to 7.0 homes, respectively. At 19 tested locations, 12 potentially explosive (Grade 1) methane concentrations of 50,000 to 500,000 ppm were detected in manholes. Financial incentives and targeted programs among companies, public utility commissions, and scientists to reduce leaks and replace old cast-iron pipes will improve consumer safety and air quality, save money, and lower greenhouse gas emissions.

A global map of urban extent from nightlights

Rural roads carry less than fifty percent of the traffic in the United States. However, more than half of the traffic accident fatalities occurred on rural roads. This research focuses on analyzing injury severities involving single-vehicle crashes on rural roads, utilizing a latent class logit (LCL) model. Similar to multinomial logit (MNL) models, the LCL model has the advantage of not restricting the coefficients of each explanatory variable in different severity functions to be the same, making it possible to identify the impacts of the same explanatory variable on different injury outcomes. In addition, its unique model structure allows the LCL model to better address issues pertinent to the independence from irrelevant alternatives (IIA) property. A MNL model is also included as the benchmark simply because of its popularity in injury severity modeling. The model fitting results of the MNL and LCL models are presented and discussed. Key injury severity impact factors are identified for rural single-vehicle crashes. Also, a comparison of the model fitting, analysis marginal effects, and prediction performance of the MNL and LCL models are conducted, suggesting that the LCL model may be another viable modeling alternative for crash-severity analysis.

Biophysical forcings of land‐use changes from potential forestry activities in North America

Urbanization, a major driver of global change, profoundly impacts our physical and social world, for example, altering not just water and carbon cycling, biodiversity, and climate, but also demography, public health, and economy. Understanding these consequences for better scientific insights and effective decision-making unarguably requires accurate information on urban extent and its spatial distributions. We developed a method to map the urban extent from the defense meteorological satellite program/operational linescan system nighttime stable-light data at the global level and created a new global 1 km urban extent map for the year 2000. Our map shows that globally, urban is about 0.5% of total land area but ranges widely at the regional level, from 0.1% in Oceania to 2.3% in Europe. At the country level, urbanized land varies from about 0.01 to 10%, but is lower than 1% for most (70%) countries. Urbanization follows land mass distribution, as anticipated, with the highest concentration between 30° N and 45° N latitude and the largest longitudinal peak around 80° W. Based on a sensitivity analysis and comparison with other global urban area products, we found that our global product of urban areas provides a reliable estimate of global urban areas and offers the potential for producing a time-series of urban area maps for temporal dynamics analyses.

Gaussian Processes for Short-Term Traffic Volume Forecasting

Land-use changes through forestry and other activities alter not just carbon storage, but biophysical properties, including albedo, surface roughness, and canopy conductance, all of which affect temperature. This study assessed the biophysical forcings and climatic impact of vegetation replacement across North America by comparing satellite-derived albedo, land surface temperature (LST), and evapotranspiration (ET) between adjacent vegetation types. We calculated radiative forcings (RF) for potential local conversions from croplands (CRO) or grasslands (GRA) to evergreen needleleaf (ENF) or deciduous broadleaf (DBF) forests. Forests generally had lower albedo than adjacent grasslands or croplands, particularly in locations with snow. They also had warmer nighttime LST, cooler daily and daytime LST in warm seasons, and smaller daily LST ranges. Darker forest surfaces induced positive RFs, dampening the cooling effect of carbon sequestration. The mean (±SD) albedo-induced RFs for each land conversion were e...

Bayesian Learning with Gaussian Processes for Supervised Classification of Hyperspectral Data

The accurate modeling and forecasting of traffic flow data such as volume and travel time are critical to intelligent transportation systems. Many forecasting models have been developed for this purpose since the 1970s. Recently kernel-based machine learning methods such as support vector machines (SVMs) have gained special attention in traffic flow modeling and other time series analyses because of their outstanding generalization capability and superior nonlinear approximation. In this study, a novel kernel-based machine learning method, the Gaussian processes (GPs) model, was proposed to perform short-term traffic flow forecasting. This GP model was evaluated and compared with SVMs and autoregressive integrated moving average (ARIMA) models based on four sets of traffic volume data collected from three interstate highways in Seattle, Washington. The comparative results showed that the GP and SVM models consistently outperformed the ARIMA model. This study also showed that because the GP model is formulated in a full Bayesian framework, it can allow for explicit probabilistic interpretation of forecasting outputs. This capacity gives the GP an advantage over SVMs to model and forecast traffic flow.