Search

The deployment of hydrogen refueling stations is vital to the development of fuel cell vehicles. Past studies on the hydrogen refueling stations didn't consider congestion on paths and assumed all flows are on the shortest paths of OD pairs, which is not consistent with travelers' route choice behavior. In this paper, we study flow refueling location allocation based on the user-equilibrium (UE) status, which is consistent with travelers' route choice behavior. We present a flow-capturing location allocation (FCLA) model on UE conditions. We also presented an equilibrium algorithm avoiding cannibalization to solve the model. Based on the transportation network of Anaheim city in California in the US, we compared the results of applying a non-equilibrium method to the FCLA model and applying the equilibrium method to the FCLA model on UE conditions. The results of the two algorithms and models are different. The equilibrium model and algorithm can capture more flow and has a higher flow-capturing ratio. What's more, the equilibrium model and algorithm can fully capture the expected flow in practice since the real transportation network is basically in UE status. The equilibrium model and algorithm presented in this paper is the first study of this kind to locate stations in cities and is useful in the deployment of hydrogen refueling stations. • Study flow refueling location allocation based on the user-equilibrium (UE) status. • A flow capturing location allocation (FCLA) model on UE conditions. • An equilibrium algorithm avoiding cannibalization to solve the model. • Compare between non-equilibrium method and equilibrium method in flow capturing. [ABSTRACT FROM AUTHOR]

Despite the good results that have been achieved in unimodal segmentation, the inherent limitations of individual data increase the difficulty of achieving breakthroughs in performance. For that reason, multi-modal learning is increasingly being explored within the field of remote sensing. The present multi-modal methods usually map high-dimensional features to low-dimensional spaces as a preprocess before feature extraction to address the nonnegligible domain gap, which inevitably leads to information loss. To address this issue, in this paper we present our novel I mbalance K nowledge- D riven Multi-modal Net work (IKD-Net) to extract features from multi-modal heterogeneous data of aerial images and LiDAR directly. IKD-Net is capable of mining imbalance information across modalities while utilizing a strong modal to drive the feature map refinement of the weaker ones in the global and categorical perspectives by way of two sophisticated plug-and-play modules: the G lobal K nowledge- G uided (GKG) and C lass K nowledge- G uided (CKG) gated modules. The whole network then is optimized using a joint loss function. While we were developing IKD-Net, we also established a new dataset called the N ational Agriculture Imagery Program and 3 D Elevation Program C ombined dataset in California (N3C-California) , which provides a particular benchmark for multi-modal joint segmentation tasks. In our experiments, IKD-Net outperformed the benchmarks and state-of-the-art methods both in the N3C-California and the small-scale ISPRS Vaihingen dataset. IKD-Net has been ranked first on the real-time leaderboard for the GRSS DFC 2018 challenge evaluation until this paper's submission. Our code and N3C-California dataset are available at https://github.com/wymqqq/IKDNet-pytorch. [ABSTRACT FROM AUTHOR]

Transportation agencies rely on scientific information to design and site infrastructure, plan operations, and make policies. However, the ways in which science interfaces with transportation policy are not straightforward: political incentives, usability challenges, implementation failures, and capacity limitations all shape how public organizations use scientific information. To build stronger connections between transportation-relevant science and policy it is necessary to better understand current patterns of scientific information use in practice, particularly as transportation agencies address an expanding purview of policy issues, such as climate change mitigation and intelligent transportation systems. This paper measures science use observed in transportation planning and project documents such as plans, proposals, impact assessments, and other deliverables that agencies are required to produce as part of the policy process. Using state, regional, and county transportation agencies in the state of California as a focal sample, this paper applies automated natural language processing (NLP) tools to analyze documents and identify references to scientific information. Our sample includes 5080 documents from 59 organizations involved in transportation governance in the state of California. We observe that the transportation science-policy interface is a cyclical system in which public agencies act in varying (and simultaneous) capacities as consumers, producers, funders, and brokers of scientific information. For instance, documents produced by state-level entities draw more heavily on academic literature, while regional and county-level agencies rely most heavily on reports produced internally by state agencies and by state-funded research institutes. Understanding where different transportation agencies access scientific information, and how scientific information flows between entities at different levels of government, can help researchers and science-policy boundary organizations increase the uptake of scientific products and design interventions to improve information access and use. • State-level transportation entities reference academic science most frequently. • Regional and county transportation organizations rely more heavily on grey literature. • Agencies draw on a broader range of scientific fields for references to emerging issues. • Transportation organizations act jointly as producers, funders, and users of science. [ABSTRACT FROM AUTHOR]

Inside orchards, turbulent coherent structures dominate the transport of heat, momentum, and moisture between the canopy and the atmosphere. Integrated quadrant analysis is a method to visualize the trajectory of individual turbulent coherent structures using in situ data from three-dimensional anemometry. In this paper, integrated quadrant analysis is used to characterize the turbulent transport of heat and momentum from two orchard experiments: one in the interrow space (the Canopy Horizontal Array Turbulence Study from Dixon, California in May and June 2007) and one in the crown of a tree (the Vertical Array Cherry Experiment from Linden, California in November 2019). By using the integrated quadrant analysis (IQA) method, this paper demonstrates the importance of the cross-wind velocity component in maintaining the turbulent coherent structures. Results from integrated quadrant analysis in three dimensions support the idea that the microfront is collocated with the boundary of a sweep and an ejection in a convective boundary layer. Moreover, in both orchards, there are preferred planar trajectories for individual coherent structures that do not depend on wind regimes. The statistical profile of the turbulence quantities, as well as individual coherent structures, are not appreciably different in the interrow space or within the crown. • IQA provides the 3D coherent structure trajectory in two orchard experiments. • Individual coherent structures show characteristic planar trajectories. • Turbulent coherent structures have peak intensity near the top of the canopy. • The cross-stream velocity is large during the passage of coherent structures. [ABSTRACT FROM AUTHOR]

• This paper proposes a comprehensive optimization program to increase economic efficiency and improve the resiliency. • A demand response program integrated with home energy storage systems is presented to optimize energy consumption. • To modify the consumption pattern of household consumers, a real-time pricing algorithm is proposed. • A self-healing system reconfiguration program integrated with distributed energy resources is presented. • Real data of California households are considered to model the home appliances and home energy storage systems. This paper proposes a comprehensive optimization program to increase economic efficiency and improve the resiliency of the Distribution Network (DN). A Demand Response Program (DRP) integrated with Home Energy Storage Systems (HESSs) is presented to optimize the energy consumption of household consumers. Each consumer implements a Smart Home Energy Management System (SHEMS) to optimize their energy consumption according to their desired comfort and preferences. To modify the consumption pattern of household consumers, a Real-Time Pricing (RTP) algorithm is proposed to reflect the energy price of the wholesale market to the retail market and consumers. In addition, a Self-Healing System Reconfiguration (SHSR) program integrated with Distributed Energy Resources (DER), reactive power compensation equipment, and Energy Storage Systems (ESSs) is presented to manage the DN energy and restore the network loads in disruptive events. The reconfiguration operation is performed by converting the isolated part of the DN from the upstream network to several self-sufficient networked virtual microgrids without executing any switching process. Real data of California households are considered to model the home appliances and HESSs. The proposed comprehensive program is validated on the modified IEEE 123-bus feeder in normal and emergency operating conditions. [ABSTRACT FROM AUTHOR]

This paper presents the first systematic study on power control strategies for Modular-Gravity Energy Storage (M-GES), a novel, high-performance, large-scale energy storage technology with significant research and application potential. Addressing the current research gap in M-GES power control technology, we propose two corresponding compensation modes and several power control strategies for Hybrid M-GES and M-GES. The effectiveness of these strategies is validated through simulations on the MATLAB/Simulink platform, including tests using sinusoidal power and feasibility verification based on a natural California load curve. Additionally, we analyze the normal and abnormal operating conditions of M-GES. In conclusion, the characteristics of the proposed control strategies are summarized, providing a guideline for further research. [Display omitted] • Modular-gravity energy storage (M-GES) power control system studied. • Two compensation modes and four control strategies are systematically studied and validated. • The center switching method with hysteresis control (CSM-H) excels in continuous compensation mode. • The center switching method with hysteresis control and dead zone (CSM-HD) excels in overlap compensation mode. • Based on the power control system proposed in this paper, the M-GES plant performs highly in natural power systems. [ABSTRACT FROM AUTHOR]

Ambitious zero-emission vehicle (ZEV) adoption goals have been proposed to decarbonize the transportation sector, while the current market share pales in comparison. Although the distinct socio-economic characteristics of ZEV early adopters relative to mainstream car buyers are well understood, the two groups' preferences for ZEV attributes are not clear. This knowledge gap hinders the development of effective policies to achieve mass ZEV penetration goals. This paper examines consumers' preferences and willingness to pay for ZEV attributes based on 755 early adopters and 3493 mainstream consumers from the 2019 California vehicle survey data. Results show that early adopters are more sensitive to battery range, acceleration performance, home charging availability, and high occupancy vehicle lane access, while mainstream consumers attach greater importance to cost attributes (e.g., fuel and maintenance costs) and charging time. Moreover, the effects of monetary incentives are found to be significant for both groups, whereas neither early adopters nor mainstream consumers value the availability of public charging stations. The findings of this study inform targeted ZEV policymaking and marketing strategies in different adoption stages. • ZEV preferences are compared between early adopters and mainstream consumers. • Early adopters are more sensitive to battery range and acceleration performance. • Mainstream consumers attach greater importance to cost attributes and charging time. • Home charging availability significantly affects BEV utility for early adopters. • Public charging infrastructure availability plays a limited role for both groups. [ABSTRACT FROM AUTHOR]

In recent works by Yang et al. (2017a, 2017b), and Yang et al. (2019), geographical, temporal, and sequential deterministic reconciliation of hierarchical photovoltaic (PV) power generation have been considered for a simulated PV dataset in California. In the first two cases, the reconciliations were carried out in spatial and temporal domains separately. To further improve forecasting accuracy, in the third case these two reconciliation approaches were applied sequentially. During the replication of the forecasting experiment, some issues emerged about non-negativity and coherency (in space and/or in time) of the sequentially reconciled forecasts. Furthermore, while the accuracy improvement of the considered approaches over the benchmark persistence forecasts is clearly visible at any data granularity, we argue that an even better performance may be obtained by a thorough exploitation of spatio-temporal hierarchies. To this end, in this paper the spatio-temporal point forecast reconciliation approach is applied to generate non-negative, fully coherent (both in space and time) forecasts. New spatio-temporal reconciliation approaches are adopted, exploiting for the first time some relationships between two-step, iterative and simultaneous spatio-temporal reconciliation procedures. Non-negativity issues of the final reconciled forecasts are discussed and correctly dealt with in a simple and effective way. The spatio-temporal reconciliation procedures are applied to the base forecasts with forecast horizon of 1 day, of PV generated power at different time granularities (1 h to 1 day), of a geographical hierarchy consisting of 324 series along 3 levels. The normalized Root Mean Square Error (nRMSE) and the normalized Mean Bias Error are used to measure forecasting accuracy, and a statistical multiple comparison procedure is performed to rank the approaches. In addition to assuring full coherence and non-negativity of the reconciled forecasts, the results show that for the considered dataset, spatio-temporal forecast reconciliation significantly improves on the sequential procedures proposed by Yang et al. (2019), at any level of the spatial hierarchy and for any temporal granularity. For example, the forecasted hourly PV generated power by the new spatio-temporal forecast reconciliation approaches improve on the NWP 3TIER forecasts in a range from 4.7% to 18.4% in terms of nRMSE. • New approach to produce fully coherent solar forecasts both in space and time. • Complete previous discussions on deterministic solar forecast reconciliation. • Avoid negative forecasts and preserve full spatio-temporal coherence. • Significant improvement of the forecast accuracy through hierarchical forecasting. • Forecast skills over NWP hourly forecasts between 4.7% and 18.4%. [ABSTRACT FROM AUTHOR]

The evaluation of communities' vulnerability to prolonged power outages offers valuable insights for prioritizing improvements in infrastructure resilience, thereby alleviating societal consequences. This study proposes a data-driven approach aiming at developing a Social Vulnerability Index (SoVI) to prolonged power outages leveraging three county-level datasets in California including (1) demographic features, (2) power outage factors, and (3) backup power factors. Furthermore, the study conducts a sensitivity analysis on three distinct datasets under two scenarios (Scenario 1: the current SoVI in 2022, Scenario 2: the prediction of SoVI in the year 2030). The results of Scenario 1 indicate that the counties with more affected customers, the number of power outages, and less education attainment tend to be more vulnerable to power outages in 2022. Scenario 1 reveals that the number of affected customers and power outages are the primary features influencing around 29% and 18% of counties, while educational attainment, public Electric Vehicles (EVs) chargers, and homes with rooftop photovoltaic (PV) substantially impact approximately 32%, 11%, and 8% of counties, respectively. However, in Scenario 2, crucial factors affecting the anticipated SoVI in 2030 include public EV chargers, houses with rooftop PV, power outages, and adults living alone. In contrast to Scenario 1, the prevalence of adults living alone has emerged as a notable factor impacting SoVI in 2030, while both scenarios underscore the pivotal role of EV chargers in influencing SoVI concerning power outages. The proposed SoVI facilitates informed policy decisions and infrastructure improvements in energy resilience, resource allocation, and disaster preparedness, contributing valuable insights for targeted interventions in these domains. • A data-driven approach was proposed to quantify the social vulnerability index (SoVI) to power outages. • This paper applies Principal Component Analysis (PCA) to identify the most correlated features from county-level datasets. • The paper conducts a sensitivity analysis to predict SoVI in2030 and evaluate how counties' dynamics change over time. • This paper provides insights into how policymakers can reduce the social vulnerability to power outages in California. [ABSTRACT FROM AUTHOR]

• The California mandate for solar PV systems on residential rooftops will have a negative occupational safety impact. • Results indicate that 16.6 recordable non-fatal injuries are expected to occur between 2020 and 2029 as a result of this policy. • The impact could be reduced by 18% by eliminating work on roofs. Introduction: A 2018 change to the California building code mandates that new residential construction in the state include rooftop solar photovoltaic power systems beginning in 2020. As residential construction (especially work on rooftops) is among the more dangerous occupations in the United States, this paper seeks to quantify the increased risks to workers as a result of this mandate. Method: An analysis of the trends by occupation of nonfatal safety incident rates in the United States combined with a Monte Carlo simulation provide an estimate of the uncertain impact of this new mandate. Results: Recordable safety incidents are anticipated to increase by a total of 16.6 incidents (standard deviation = 1.0 incidents) over the 2020–2029 time period as a result of this policy change. However, lessons from Germany and other industries offer potential avenues to reduce the negative social impact of this mandate. Conclusions: While it is not possible to increase employment in any sector without increasing the expected number of occupational injuries to some degree, these results indicate that risks could be considerably reduced by making solar PV system design decisions that increase worker productivity and reduce roof exposure time. Practical Applications: Changes such as eliminating work on roofs could decrease the expected number of recordable injuries over the 10-year period by 0.30 incidents per year (a reduction of 18%). [ABSTRACT FROM AUTHOR]

For renewable hydrogen to be a significant part of the future decarbonized energy and transportation sectors, a rapid and massive build-out of hydrogen production facilities will be needed. This paper describes a geospatial modeling approach to identifying the optimal locations for renewable hydrogen fuel production throughout the state of California, based on least-cost generation and transport. This is accomplished by (1) estimating and projecting California renewable hydrogen demand scenarios through the year 2050, (2) identifying feedstock locations, (3) excluding areas not suitable for development, and (4) selecting optimal site locations using commercial geospatial modeling software. The findings indicate that there is a need for hundreds of new renewable hydrogen production facilities in the decades preceding the year 2050. In selecting sites for development, feedstock availability by technology type is the driving factor. • Extensive build-out of new renewable hydrogen facilities is needed for hydrogen economy. • Zoning and terrain limit the potential renewable hydrogen production facility sites. • Proximity to feedstock dominates cost-optimal site selection. [ABSTRACT FROM AUTHOR]

Many nations and jurisdictions have legalized non-medical adult use of cannabis, or are considering doing so. This paper contributes to knowledge of adult use legalization's associations with cannabis use disorder (CUD) treatment utilization. This study collected data from a dataset of all publicly funded substance use disorder treatment delivered in California from 2010 to 2021 (1,460,066 episodes). A logistic regression model estimates adult use legalization's impacts on CUD treatment utilization using an individual-level pre-post time series model, including individual and county-level characteristics and county and year-fixed effects. Adult use legalization was associated with a significant decrease in the probability of admission to CUD treatment (average marginal effect (AME): −0.005, 95 % CI: −0.009, 0.000). Adult use legalization was also associated with a decrease in the probability of admission to CUD treatment for males (AME: −0.025, 95 % CI: −0.027, −0.023) Medi-Cal beneficiaries (AME: −0.025, 95 % CI: −0.027, −0.023) adults ages 21+ (AME: −0.011, 95 % CI: −0.014, −0.009) and Whites (AME: −0.012, 95 % CI: −0.015, −0.010), and an increase in the probability of admission to CUD treatment for patients referred from the criminal justice system (AME: 0.017, 95 % CI: 0.015, 0.020) and Blacks (AME: 0.004, 95 % CI: 0.000, 0.007) and Hispanics (AME: 0.009, 95 % CI: 0.006, 0.011). Adult use legalization is associated with declining CUD treatment admissions, even though cannabis-related problems are becoming more prevalent. Policies and practices that protect public health, and engage people with CUD in treatment are needed. • Pre-legalization, 19 % of SUD admissions were CUD, compared to 9 % post-legalization. • Adult use legalization was associated with a decrease in CUD treatment admissions. • Differences existed in legalization's associations on different demographic groups. [ABSTRACT FROM AUTHOR]

Liquefaction hazard analyses are needed for regional-scale applications, such as risk analysis and upgrades of electrical transmission, water conveyance, and highway systems. In this paper, we expand well-established site-specific liquefaction triggering and manifestation analysis methods for regional scale application by explicitly addressing the uncertainty in groundwater conditions, soil behavior classification, and cyclic resistance. The uncertainty inherent to regional-scale geotechnical properties is quantified by considering the inputs into conventional liquefaction triggering and manifestation analyses as random variables. With this flexible and extensional formulation, the geostatistical properties needed for regional scale analyses may be estimated by either averaging borehole information or spatially interpolating geotechnical properties across broad geologic deposits. We demonstrate the utility of the method by analyzing the liquefaction hazard throughout the south San Francisco Bay area, California, U.S.A. The resulting estimates of the probability of liquefaction manifestation tend to be lower than previously published estimates, largely owing to the geologic structure of the area and the uncertainty in liquefaction surface manifestation once liquefaction is triggered at depth. Uncertainties in groundwater conditions, soil behavior classification, penetration resistance, and liquefaction manifestation are quantified by the regional scale analyses, which provide a complete and systematic evaluation of the liquefaction hazards, localize the uncertainties in subsurface conditions, and facilitate supplemental investigations at areas of high uncertainty to fill data gaps. • Existing liquefaction models are extended for regional-scale application. • Flexible model allows deposit-scale averages or spatially interpolated properties. • Geologic structure and uncertain groundwater depth are applied in the model. • Example application in southern San Francisco Bay area, California, U.S.A. [ABSTRACT FROM AUTHOR]

The quality of traffic services depends on the accuracy and completeness of the collected traffic data. However,the existing traffic data imputation methods usually only rely on the predefined road network structure to capture the spatio-temporal features and only consider the imputation effect from a single perspective, which are very limited for imputation of different missing patterns of road traffic data. In this paper, we propose a novel deep learning framework called Hierarchical Spatio-temporal Graph Convolutional Neural Networks(HSTGCN) to impute traffic data,through the macro layer and the road layer. The model constructs macro graph of the road network based on the data temporal correlation clustering, which can mine the temporal dependencies of road traffic data from a hierarchical perspective. Besides, a temporal attention mechanism and adaptive adjacency matrix are introduced in the road layer to better extract the spatio-temporal information of the road traffic data. Finally, we use graph convolution neural networks to learn the spatio-temporal feature representations of the road layer and macro layer, which are then fused to achieve data imputation. To illustrate the efficient performance of the model, experiments are conducted on traffic data collected from California and Seattle. The proposed model performs better than the comparison model for traffic data imputation. • A full dynamic graph is built based on the urban road structure. • Extracting spatio-temporal features from multiple dimensions using hierarchical thinking. • Gate-GCN perform effective feature selection. [ABSTRACT FROM AUTHOR]

• Factors affecting injury severity in rollover crashes were explored. • Random parameters logit model with unobserved heterogeneity in means and variances. • Findings reveal temporal instability of model specifications across individual years. • Aggregate-to-component shift quantified using out-of-sample prediction. • Implications for mitigating the associated driver injury severity. Single-vehicle rollover crashes have been acknowledged as a predominant highway crash type resulting in serious casualties. To investigate the heterogeneous impact of factors determining different injury severity levels in single-vehicle rollover crashes, the random parameters logit model with unobserved heterogeneity in means and variances was employed in this paper. A five-year dataset on single-vehicle rollover crashes, gathered in California from January 1, 2013, to December 31, 2017, was utilized. Driver injury severities that were determined to be outcome variables include no injury, minor injury, and severe injury. Characteristics pertaining to the crash, driver, temporal, vehicle, roadway, and environment were acknowledged as potential determinants. The results showed that the gender indicator specified to minor injury was consistently identified as a significant random parameter in four years' models and the joint five-year model, excluding the 2016 crash model where the night indicator associated with no injury was observed to produce the random effect. Additionally, two series of likelihood ratio tests were conducted to assess the year-to-year and aggregate-to-component temporal stability of model estimation results. Marginal effects of explanatory variables were also calculated and compared to analyze the temporal stability and interpret the results. The findings revealed an overall temporal instability of model specifications across individual years, while there is no significant aggregate-to-component variation. Injury severities were observed to be stably affected by several variables, including improper turn indicator, under the influence of alcohol indicator, old driver indicator, seatbelt indicator, insurance indicator, and airbag indicator. Furthermore, the year-to-year and aggregate-to-component shift was quantified and characterized by calculating the differences in probabilities between within-sample observations and out-of-sample predictions. The overall results imply that continuing to expand and refine the model to incorporate more comprehensive datasets can result in more robust and stable injury severity prediction, thus benefiting in mitigating the associated driver injury severity. [ABSTRACT FROM AUTHOR]

Forest fires cause every year damages to biodiversity, atmosphere, and economy activities. Forest fire simulation have improved significantly, but input data describing fire scenarios are subject to high levels of uncertainty. In this work the two-stage prediction scheme is used to adjust unknown parameters. This scheme relies on an input data calibration phase, which is carried over following a genetic algorithm strategy. The calibrated inputs are then pipelined into the actual prediction phase. This two-stage prediction scheme is leveraged by the cloud computing paradigm, which enables high level of parallelism on demand, elasticity, scalability and low-cost. In this paper, all the models designed to properly allocate cloud resources to the two-stage scheme in a performance-efficient and cost-effective way are described. This Cloud-based Urgent Computing (CuCo) architecture has been tested using, as study case, an extreme wildland fire that took place in California in 2018 (Camp Fire). • Data-driven calibration to deal with uncertainty in forest fire spread prediction. • Cloud-based urgent computing implementation of a two-stage prediction model. • Use of utility function to deal with the cost-performance trade-off. • Validation against a deadly and destructive wildfire with promising results. [ABSTRACT FROM AUTHOR]

Disasters have been a long-standing concern to societies at large. With growing attention being paid to resilient communities, such concern has been brought to the forefront of resilience studies. However, there is a wide variety of definitions with respect to resilience, and a precise definition has yet to emerge. Moreover, much work to date has often focused only on the immediate response to an event, thus investigating the resilience of an area over a prolonged period of time has remained largely unexplored. To overcome these issues, we propose a novel framework utilizing network analysis and concepts from disaster science (e.g., the resilience triangle) to quantify the long-term impacts of wildfires. Taking the Mendocino Complex and Camp wildfires - the largest and most deadly wildfires in California to date, respectively - as case studies, we capture the robustness and vulnerability of communities based on human mobility data from 2018 to 2019. The results show that demographic and socioeconomic characteristics alone only partially capture community resilience, however, by leveraging human mobility data and network analysis techniques, we can enhance our understanding of resilience over space and time, providing a new lens to study disasters and their long-term impacts on society. • Quantifying community resilience is an open research challenge • This paper develops a novel framework to quantify resilience after a disaster using network analysis • Using human mobility data associated with two wildfires, we measure robustness and vulnerability of communities • Results show community resilience closely tied to socio-economic and built environmental traits of the affected areas • Our approach paves a way to study disasters and their long-term impacts on society [ABSTRACT FROM AUTHOR]

• We analyzed trends and breakpoints for three streamflow signatures of drying patterns in 540 non-perennial streams over CONUS • Nonlinear changes in streamflow are more common than linear changes for all three stream intermittency signatures. • The past behavior of non-perennial streams might be a poor predictor of future drying patterns Stream drying patterns – including duration, timing, and dry-down rates – affect aquatic ecosystems and nutrient exports in non-perennial streams. Because hydrologic processes are often nonlinear, changes in drying may also be nonlinear, but analyses of historical changes in stream drying to date have not characterized the frequency or functional forms of nonlinear change. Understanding the extent of nonlinear change in non-perennial streams is essential for advancing our fundamental knowledge of hydrological processes, aquatic ecosystems, and watershed functioning under a warming climate. This paper uses a polynomial-based trend detection technique (PolyTrend) to analyze the linear and nonlinear trend behaviors of three intermittency signatures (annual no-flow days specifying longer or shorter drying duration, day of first no-flow occurrence specifying timing of stream drying, and days from peak to no-flow specifying dry-down rates) at 540 non-perennial gage stations over 38 years (1980–2017) across the continental United States (CONUS). Additionally, we carried out a breakpoint analysis to characterize the discontinuities in the time series of each intermittency signature. Analysis of annual no-flow days shows that about 37 % of the total streamflow stations are drying for longer each year, whereas about 22 % are wetter for longer than in the past. The day of first no-flow occurrence analysis shows that 10 % of the streams are drying earlier, and 19 % are drying later. On the other hand, analysis of days from peak to no-flow shows that 14 % of streams are drying faster, and 17 % are drying more slowly. For all these metrics, among the significant trends, at least half of the relationships were nonlinear. For annual no-flow days , the breakpoint analysis shows more discontinuities in the second half of the analysis period (1999 to 2017) than in the first half, with more discontinuities in the Southern Great Plains than in other regions. The other two signatures demonstrate less frequent discontinuities in the second half of the analysis period, suggesting decreased nonlinear dynamics in recent years. Nonlinear no-flow duration trends are common in Mediterranean California, and the dry-down rate has increased in recent decades. Our findings indicate that nonlinear change in stream drying is widespread and must be accounted for in watershed planning and management. [ABSTRACT FROM AUTHOR]

• Impact of COVID-19 on ECE programs differed greatly by program type, funding source. • Family child care homes fared worse in most measures of economic well-being with direct implications for individual providers during COVID-19. • Lower attendance, staffing concerns were pandemic challenges for center-based care. • Voucher-receiving centers more likely to face negative impacts during COVID-19. • Head Start/state-contract centers more able to support staff well-being in pandemic. The COVID-19 crisis has overwhelmed and weakened the United States early care and education (ECE) sector, jeopardizing a system that was already precariously situated atop a weak foundation. While multiple national- and state-level studies have highlighted the overwhelming impacts of the pandemic on the ECE sector, little has been reported about how much variation in impacts exists, and in what forms, within the ECE sector. Based on a statewide survey of 953 licensed care providers in California conducted in June 2020, this paper examines the impact of COVID-19 experienced by ECE providers, focusing on the variations between centers and family child care homes (FCCs) and among center-based programs. Results indicate that the challenges programs face differ greatly depending on program type and funding source. Compared to center-based programs, FCCs fared worse in most measures of economic hardship that directly impact individual providers with medium to large effect sizes. Centers were more likely than FCCs to struggle with reduced attendance and changes in program operations by medium to large effect sizes and report staffing challenges by small to medium effect sizes. Among the center-based programs, subsidized programs holding contracts with Head Start or the California Department of Education (such as state preschool programs) were more stable and better able to financially support their staff during the pandemic, with effect sizes ranging from medium to large. Centers receiving government subsidies in the form of vouchers were more likely to be negatively impacted by the pandemic compared to unsubsidized centers and Head Start and state-contracted centers. Implications for future research and policy are discussed in the context of addressing the complex delivery system of ECE services and supporting outcomes that are effective and equitable for children, families, and the ECE workforce. [ABSTRACT FROM AUTHOR]

An alarming, consistent increase in U.S. pedestrian fatalities since 2009 culminated in a 28-year high of 6,283 pedestrians killed in 2018. Yet these numbers obscure a second alarming trend: 75% of pedestrian fatalities occur in darkness, and nearly 90% of the increase in pedestrian fatalities from 2009 to 2018 occurred in darkness. This paper examines data on pedestrian fatalities at the national level and pedestrian fatalities and serious injuries in California from 2012 to 2017 to better understand correlates of severe pedestrian injuries in darkness. Binary and multinomial logit models reveal that variables related to roadway design and operations (e.g., speed limits, number of lanes, roadway type, and presence of traffic control) – but not speeding – are significantly associated with the likelihood of a pedestrian fatality or serious injury occurring in darkness as compared to daylight. Critically, these factors – which were consistent for fatalities regardless of lighting presence and roadway type, with few exceptions – are all worse in darkness because they are negatively affected by a lack of visibility. Alcohol usage by drivers or pedestrians and sociodemographic characteristics were also positively associated with severe injuries in darkness. Our findings urge an explicit consideration of pedestrian safety in darkness in all future design and retrofit decisions, and particularly on higher-speed, multi-lane roadways. Immediate solutions include roadway designs and policies that slow drivers, particularly at night, and that increase illumination and driver attention, such as through additional roadway lighting, high-visibility countermeasures and protected crossings, and adaptive lighting and detection technology for vehicles. • Pedestrian fatalities on multilane roads significantly more likely in darkness. • Fatalities at speeds over 30 mph are significantly more likely in darkness. • Black and Native American pedestrian deaths significantly more likely in darkness. • Efforts to slow drivers and provide safe crossings in darkness are needed. • Roadway design efforts should explicitly consider pedestrian safety in darkness. [ABSTRACT FROM AUTHOR]

• Analysis of 275-day long time series of county level COVID-19 data in California state. • Multiple change point estimation in the framework of mixture modeling and model-based clustering. • Change point given by gap between logit transformed segments in negative binomial nonhomogeneous Levy process. • Three geographically meaningful clusters, each with several change points indicating the spread and decline of infection. In this paper, a model-based clustering algorithm relying on a finite mixture of negative binomial Lévy processes is proposed. The algorithm models heterogeneous stochastic count process data and automatically estimates multiple change points upon fitting the mixture model. Such change point estimation identifies time points when deviation from the standard process has occurred and serves as an important diagnostic tool for analyzing temporal data. The proposed model is applied to the COVID-positive ICU cases in the state of California with very interesting results. [ABSTRACT FROM AUTHOR]

This paper examines the relationship between smartphone use by drivers and traffic accidents in California between 2001 and 2013. In order to estimate smartphone use, we first show that widespread adoption of modern smartphones began in 2009 after the release of the iPhone 3G and T-Mobile G1. This information is combined with annual 3G coverage maps that are constructed from cellular tower information in a machine learning framework. In a difference-in-differences framework, we estimate the combined effect of smartphone adoption and 3G coverage along quarter-mile road segments. Controlling for census tract population density, road and year fixed effects, Poisson regression results show that there is a statistically significant increase in the traffic accident rate along a road segment when smartphone use becomes possible. Our preferred specification suggests smartphones caused accident rates to increase by 2.9 percent, resulting in 3500 additional accidents per year in California. Event study results rule out the possibility that our smartphone treatment is capturing a trend in the accident rate. The results are robust to a variety of specifications and consistent with individual-level studies showing that cell phone use leads to lower driving quality. The findings also provide guidance for policies aimed at reducing cell phone related accidents and distracted driving. [ABSTRACT FROM AUTHOR]

Machine learning methods are increasingly being used in the field of air quality research to investigate the relationship between air pollutant levels, emissions, and meteorological changes over time. This research is used for both scientific investigation, and policy assessment and development. However, there is a lack of studies that have compared the performance of different machine learning methods. To address this gap, this paper employed various machine learning techniques, including decision tree, random forest (RF), support vector machine (SVM), support vector regression (SVR), k-nearest neighbor, neural network, and Gaussian process regression, to predict daily average PM 2.5 levels and the number of days with PM 2.5 exceedance in the South Coast Air Basin of California from 2000 to 2019. The models were trained using meteorological factors, estimated emissions, and large-scale climate indices as inputs. The SVR model demonstrated the highest predictive accuracy for PM 2.5 levels and the SVM model gave the most accurate results for predicting the number of days with PM 2.5 exceedances. Conversely, the decision tree model performed the least accurately. The results also showed that emissions have a greater impact on PM 2.5 levels over time compared to meteorological factors, though meteorology is responsible for daily variability. The most important meteorological factors were identified as surface relative humidity and relative humidity at 850 mbars, which are related to partitioning, cloud cover and wet deposition. We conducted sensitivity tests on the model's response to emissions and meteorological factors. The predicted PM 2.5 from RF and SVR showed large correlations with emissions at the early period (2000–2010). However, the changes were minimal in more recent years (2011–2019), implying that there are biases in machine learning models, in which the models consistently predict the minimum PM 2.5 levels at a baseline. [Display omitted] • SVR is more accurate on daily PM 2.5 , particularly PM 2.5 exceedances predictions. • Surface RH was the most important meteorological factor for PM 2.5 prediction. • The impact of emissions on PM 2.5 was significant before 2010 but reduced thereafter. • ML models predict past better than future; all the ML models are limited at extremes. [ABSTRACT FROM AUTHOR]

How much should electric utilities pay to maintain a reliable electricity system? This paper describes an open-ended stated-preference experiment that generates estimates for how advanced notification impacts household willingness-to-pay (WTP) to avoid outages. We find positive and statistically significant WTP to avoid power outages of $10/kWh, consistent with the expectation that outages are costly to the residential sector. We find notification reduces the WTP, but the effects are not statistically significant. There is limited evidence that these results vary by income and wealth levels. Back-up power ownership is positively correlated with respondents' WTP to avoid outages. • Stated preference survey estimates WTP to avoid power outages. • Notifications reduce WTP but result is not statistically significant. • Back-up power ownership is positively correlated with WTP. • Results inform deployment of California PSPS and EPSS programs. [ABSTRACT FROM AUTHOR]

Inequities in the built environment have long persisted in communities around the world. Prior studies have exposed that energy retrofit programs invest inequitably in residential buildings, leaving racial minorities with disproportionately less energy inefficient homes, regardless of income level. This inequity in energy performance and retrofit investments has yet to be studied for commercial and institutional buildings. Further research has shown that tree canopy coverage and vegetation is less dense in historically disadvantaged communities, contributing to a greater urban heat island effect and corresponding higher energy consumption for cooling compared to historically advantaged communities. Moreover, there is limited albeit emerging research that utilizes granular socioeconomic data with fine temporal scale energy consumption data to help identify energy inequalities among race and income groups. This paper examines the differential impact of energy efficiency retrofit installation on energy performance between advantaged vs. disadvantaged groups using data from U.S. public school buildings. Additionally, it studies how the effect of retrofits is compounded by the presence of vegetation surrounding the building. Utilizing hourly interval smart meter data from primary and secondary school buildings in California, the impact of retrofits are quantified using three energy metrics. Then the disparities in retrofit impact for advantaged vs. disadvantaged groups (as defined by three equity metrics) are evaluated using school-level demographic data. Lastly, regression on the percent change in monthly energy consumption from the pre-retrofit to post-retrofit period is used to quantify the incremental benefits of retrofits compounded by vegetation. Findings show that energy efficiency retrofits can improve energy performance and reduce energy inequities. Specifically, retrofits significantly increase energy efficiency and reduce the energy efficiency equity gap by up to 198%. Evidence shows that vegetation surrounding a building has a synergistic effect with mechanical retrofits for reducing energy consumption in respect to reduced cooling loads, even in cases where mechanical retrofits alone resulted in significant increases in energy consumption. Dense vegetation provides around 6% energy savings, similar in magnitude to savings from mechanical retrofit installation. This validates the effectiveness of retrofits for reducing energy consumption and improving energy equity, and demonstrates the ability of vegetation to provide additional energy savings beyond mechanical interventions. [ABSTRACT FROM AUTHOR]

To assess whether a measure of leadership support for worker safety, health, and well-being predicts staff turnover in nursing homes after controlling for other factors. This paper uses administrative payroll data to measure facility-level turnover and uses a survey measure of nursing home leadership commitment to workers. In addition, we use data from Medicare to measure various nursing home characteristics. Nursing homes with at least 30 beds serving adults in California, Ohio, and Massachusetts were invited to participate in the survey. The analysis sample included 495 nursing homes. We used a multivariable ordinary least squares model with turnover rate as the dependent variable. We used an indicator for nursing homes who scored above the median on the measure of leadership that supports worker safety, health, and well-being. Control variables include bed count (deciles), ownership (corporate/noncorporate × for-profit/not-for-profit), percent of residents on Medicaid, state, being in a nonmetropolitan county, and total nurse staffing per patient day in the 2 quarters before the survey. The unadjusted turnover rate was lower for those nursing homes that scored higher on leadership commitment to worker safety, health, and well-being. After controlling for additional variables, greater leadership commitment was still associated with lower turnover but with some attenuation. We find that nursing homes with leadership that communicated and demonstrated commitment to worker safety, health, and well-being had relatively fewer nurses leave during the study period, with turnover rates approximately 10% lower than homes without. These findings suggest that leadership may be a valuable tool for reducing staff turnover. [ABSTRACT FROM AUTHOR]

Nitrate pollution of surface water from farms and urban runoff is widespread – impairing drinking water supplies, recreation, and wildlife habitat. The scale of the problem in rivers has overwhelmed most watershed-based best management practices. A new paradigm is needed. In this paper we highlight the 119 ha (294 acres) Prado Wetlands in southern California, USA as an example of a gravity-enabled, free surface flow vegetated wetlands to treat a river – the nitrate-rich, effluent-dominated Santa Anna River. The wetland treats ∼1–2 m3/s of the 7.1 m3/s (250 cfs) flow of the river in 28 wetland cells in two parallel trains (residence time 2–4 d, hydraulic loading rate 10–20 cm/d). The cleaner water is returned to the river and infiltrated downstream into a managed groundwater basin supplying 2.4 million customers. Since 1998, ∼1200 t of nitrate-N have been permanently removed from the river and released as N 2 -gas into the air. Over 24 years, in the warm summers an average of 6.1 mg-N/L entered the wetland and 3.7 mg-N/L flowed out, a removal efficiency of 57%. In the cooler winters, an average of 6.4 mg NO 3 -N/L entered and 2.6 mg NO 3 -N/L flowed out, a removal efficiency of 45%. Individual months varied widely in concentration removal efficiency from 7% (winter) to 98% (high summer). Areal summer removal averaged 137 mg NO 3 -N/m2/d and the apparent "k" was around 46 m/y. All-year rates were 129 mg NO 3 -N/m2/d or 47 g/m2/y, a reasonable performance considering the size of the system and its configuration for both nitrate removal and duck hunting (i.e., more open water). Annual average nitrate removal efficiency increased substantially over time (45 to 72% for 1998–2004 versus 2016–22) and is attributed to an increase in hydraulic residence time, upgraded internal plumbing, and increased winter temperatures (+3C°). Prado Wetlands also removes phosphorus, though not as well as nitrate; average warm season removal efficiency was 23% (inflow 1.20 mg PO 4 -P/L, outflow 0.93 mg PO 4 -P/L). The aerial removal rate was high at 23 mg P/m2/d or 8.5 g/y, probably due to the high inflowing P-concentrations and the hard water. Wildlife was also enhanced. Least Bell's Vireo (Vireo belli pusillus) , an endangered bird in the Prado Basin, increased from 19 pairs (1998) to a recent average of 444 pairs due to habitat improvement and reduction of cowbird parasitism. Results from the Prado Wetlands show that cleaning up rivers via diversion through wetlands benefits both water quality and wildlife. [ABSTRACT FROM AUTHOR]

• Experimental performance of 45 m2 non-tracking solar thermal array described. • Performance evaluated at 80 °C, 135 °C, and 170 °C in dirty and clean conditions. • Annual thermal generation estimated 1000 kWh/m2 at 100 °C and 700 kWh/m2 at 160 °C. • $300/m2 installed, $6.5/m2 annual O&M, levelized cost of heat 2–4 cents per kWh. • Applications in desalination, evaporation, steam generation, process heating. In this paper, the experimental performance of a 45 m2 solar field of non-tracking external compound parabolic (XCPC) collectors installed at the University of California, Merced is described. The solar field was operated during July-August 2020 in both clean and dirty conditions and at varying operating temperatures (70, 135, 170 °C) while operating an air heater, thermal evaporator, and double effect absorption chiller. Performance data was used to develop an instantaneous solar field performance model which was then incorporated into an annual performance model using TMY3 data to estimate yearly production from the solar field. The model predicts an annual generation of ∼1100 kWh/m2-year at 80 °C, ∼1000 kWh/m2-year at 100 °C, ∼900 kWh/m2-year at 120 °C, ∼800 kWh/m2-year at 140 °C, and ∼700 kWh/m2-year at 160 °C in California. The XCPC technology is currently expected to have an installed cost of $300/m2 and an annual operations and maintenance cost of $6.5/m2-year. Over a 25 year lifetime it provides a levelized cost of heat at 2–4 cents per kWh th delivered. This is below the cost of commercial natural gas in California and at temperatures ≤ 120 °C below the cost of industrial natural gas, which highlights the potential of the XCPC technology for decarbonizing thermal applications such as water and space heating, drying, sterilization, desalination, evaporation, low pressure steam, double effect absorption chilling, process heating, and more. The lifetime cost of emissions reductions is ∼$169 per metric ton of avoided CO 2 when replacing natural gas, ∼$137/MT CO 2 when replacing propane, and ∼$83/MT CO 2 when replacing electric heating. [ABSTRACT FROM AUTHOR]

Understanding how maritime hunter-gatherer diets changed through time in response to increasing social complexity can help us understand important transitions in early human history. This paper presents new baseline stable isotope values for southern California with an emphasis on marine plant and animal species. We use our baseline database to reevaluate human stable isotope values from the region using Bayesian mixing models to interpret dietary patterns across time and geographic space. Our analysis compares categories of foods consumed between island, coastal, and interior populations across the Middle and Late Holocene (circa 8000 to 168 cal BP) occupational history of precolonial southern California. Our results show a clear increase in the importance of high trophic marine foods, such as finfish, relative to low trophic level food, such as shellfish through time, paralleling increases in population size, economic intensification, and village aggregation in the Channel Region. This case study displays the capacity of Bayesian modeling to infer patterns of dietary change in the past when applied to human isotope values and adds to previous studies on the relationship between population growth, technological innovation, and the intensification of resource extraction in the region. [ABSTRACT FROM AUTHOR]

With a recently concluded pilot program in the State of California, road user charge (RUC) was explored as an alternative means of transportation funding. The goal was to gather public opinion, address privacy and data security concerns. The pilot program was meant for the future implementation of RUC in the State. However, the impact of RUC on travel demand changes impacting California's industry sectors was not explored or probed. It is expected that by 2020 or later, RUC could have the potential to be implemented in the State causing some initial mode shifts or reduced number of trips made by private vehicles to minimize the cost of travel. This could have a potential impact on employment-based trips to industry sectors of California. Therefore, this paper explores the effects of RUC on the 'attractiveness' of four key industry sectors of the State using potential accessibility changes. The changes are evaluated from the RUC implementation for the year 2019 using the year 2013 as the base year. The four key industry sectors used in the evaluation are (i) Manufacturing, (ii) Retail Trade, (iii) Health Care and Social Assistance, and (iv) Accommodation and Food Services. In this evaluation, four vehicle fuel-types are considered, namely gasoline, compressed natural gas (CNG), diesel, and electric for California's rural and urban counties. On average, the percentage change in potential accessibility was found to be largest for counties for the Health Care and Social Assistance industry sector across all fuel types for both rural and urban counties in California. This indicates that the commuters from this industry sector would experience the most considerable benefit with RUC implementation. The lowest average percentage potential accessibility change was observed for the Accommodation and Food Services. This sector's labor force would be the least affected by the RUC implementation for the four industry sectors analyzed. • Accessibility impacts due to road user charge (RUC) in California are evaluated. • Impacts are studied for four key industry sectors of California. • Vehicle fuel types are also taken into consideration. • Analyses are carried out for RUC implementation in California. • Health Care and Social Assistance industry commuters will have the largest accessibility benefit. [ABSTRACT FROM AUTHOR]

This paper offers a comparative look at the organization of built space and storage for sedentary communities in the world's Mediterranean climatic regions and assesses the variable expressions of certain underlying features that are held in common. Our analysis focuses on two primary case studies: the occupants of the Aegean Sea region from the Neolithic to Roman periods (ca. 6800 BCE-168 CE) and the coastal Chumash of southern California during the Late Period (1300–1782 CE). We show that in both regions, households took advantage of outdoor spaces as productive activity areas during the long periods of favorable weather each year. This incorporation of exterior spaces was typically structured using clear visual cues or architecture. Additionally, surplus food was acquired by individual households and stored to mitigate the inherent risks of interannual variability in precipitation and expected annual periods of lower yield (i.e., winter). Since storage was typically organized at the household level, the domestic architecture often incorporated dedicated space or built facilities to manage this surplus. By understanding such climatic influences on the regions' architecture and storage infrastructure, the distinct cultural qualities of these expressions can be better articulated. While it is not our contention that these attributes manifested identical outward architectural and material forms or were exclusive to Mediterranean environments, we suggest that the environment was an important influence on construction and storage decisions which could be adapted according to the local histories of their respective communities. [ABSTRACT FROM AUTHOR]

This study characterized nitrogen oxides (NOx) and ammonia (NH 3) emissions from five in-use goods movement vehicles, including one diesel vehicle without selective catalytic reduction (SCR), two diesel vehicles with SCR, and two ultra-low NOx natural gas vehicles equipped with three-way catalysts (TWCs). Emissions testing was performed under real-world driving conditions in typical goods movement routes in the South Coast Air Basin (SCAB) of California. NOx emissions varied depending on the vehicle and the route. The no-SCR diesel vehicle showed the highest NOx emissions over all the routes. One of the SCR-equipped vehicles showed NOx emission rates that were two times higher than the other SCR-equipped vehicle with similar engine model year and mileage, which may be attributed to catalyst deterioration. The SCR-equipped diesel vehicles were within two to three times 0.2 g/bhp-hr certification standard over the different routes. The natural gas vehicles showed average NOx emissions around or below the optional Low NOx emissions standard of 0.02 g/bhp-hr. The three-bin moving average window (MAW) method was utilized to show the NOx emissions across various modes of operation, including idle, low load, and medium/high load. The highest fraction of vehicle operation was found in either the low load or medium/high load bins, depending on the route. NOx emissions for each of the bins varied between different vehicles, and are further analyzed in this paper. NH 3 emissions formation was favored for the TWC-equipped natural gas vehicles, which produced about 20 times more NH 3 emissions than the SCR-equipped diesel vehicles. [Display omitted] • Significant NOx reductions for the natural gas vehicles compared to the SCR-equipped diesel vehicles. • NH 3 formation favored in the TWC for the natural gas vehicles and produced during acceleration events. • Low load and medium/high load conditions dominated vehicle operation and NOx emissions. [ABSTRACT FROM AUTHOR]

Vegetation structure mapping is critical for understanding the global carbon cycle and monitoring nature-based approaches to climate adaptation and mitigation. Repeated measurements of these data allow for the observation of deforestation or degradation of existing forests, natural forest regeneration, and the implementation of sustainable agricultural practices like agroforestry. Assessments of tree canopy height and crown projected area at a high spatial resolution are also important for monitoring carbon fluxes and assessing tree-based land uses, since forest structures can be highly spatially heterogeneous, especially in agroforestry systems. Very high resolution satellite imagery (less than one meter (1 m) Ground Sample Distance) makes it possible to extract information at the tree level while allowing monitoring at a very large scale. This paper presents the first high-resolution canopy height map concurrently produced for multiple sub-national jurisdictions. Specifically, we produce very high resolution canopy height maps for the states of California and São Paulo, a significant improvement in resolution over the ten meter (10 m) resolution of previous Sentinel / GEDI based worldwide maps of canopy height. The maps are generated by the extraction of features from a self-supervised model trained on Maxar imagery from 2017 to 2020, and the training of a dense prediction decoder against aerial lidar maps. We also introduce a post-processing step using a convolutional network trained on GEDI observations. We evaluate the proposed maps with set-aside validation lidar data as well as by comparing with other remotely sensed maps and field-collected data, and find our model produces an average Mean Absolute Error (MAE) of 2.8 m and Mean Error (ME) of 0.6 m. [Display omitted] • Very high resolution canopy height maps at jurisdictional scale are released. • Improved performance from vision transformers based on Self-Supervised Learning (SSL). • First use of SSL and vision transformers for canopy height estimation. • Low resolution GEDI and high resolution aerial lidar predictions are combined. • Model generalizes well to aerial imagery, even though trained with satellite images. [ABSTRACT FROM AUTHOR]

• RCA-clay mixture can be a reliable and eco-friendly alternative as subgrade layers. • Recycled concrete aggregate (RCA) inclusion in expansive clay subgrade leads to higher strength and resilience modulus. • The optimum proportion of Clay-20 %RCA-9 %GBS mixture is found to be strong enough to be used as subgrade/subbase layers. • The model proposed by MPDGE provides a good prediction for the M r values of treated expansive clay subgrade. • This paper can encourage and further promote the use of recycled concrete aggregate in the road construction industry. In urban areas, due to urban development and lack of space, roads may be built on the expansive clay subgrade. Expansive clay soils increase in volume when in contact with water and shrink by losing water. Recycled concrete aggregates (RCA) are one of the components of construction and demolition (C&D) wastes that can be used instead of natural resources in road construction projects. In this research, a combined method of mechanical and chemical stabilization was used to evaluate the use of RCA as a mechanical stabilizer and granulated blast furnace slag (GBS) as a chemical stabilizer to improve the strength and resilience characteristics of expansive clay subgrade. Laboratory results showed that adding RCA wastes to expansive clay subgrade reduced atterberg limits (LL, PL, and PI) and optimum water content (OWC) and increased the maximum dry density (MDD), California bearing ratio (CBR), and uniaxial compression strength (UCS). Expansive clay subgrade with 20 % RCA was treated with GBS chemical stabilizer in order to achieve a minimum 7 days target strength of 1.7 MPa for subbase application. A series of repeated load triaxial (RLT) tests were performed according to the AASHTO T307 procedure and resilient modulus (M r) values of the stabilized clay subgrade specimen were evaluated at different curing times under cyclic loading. Also, the proposed Mechanistic-Empirical pavement design Guide (MEPDG) model was used to predict M r values in different confining and deviator stresses which showed a good correlation with the measured M r values. Finally, the improvement of the clay subgrade stabilized with RCA and GBS was also supported by SEM analysis. The results obtained from this research showed that the use of RCA, in addition to reducing disposed waste and environmental pollution, can improve the physical and mechanical characteristics of expansive clay Subgrade layers. [ABSTRACT FROM AUTHOR]

Mandated minimum nurse-to-patient ratios have been the subject of active debate in the U.S. for over twenty years and are under legislative consideration today in several states and at the federal level. This paper uses the 1999 California nurse staffing mandate as an empirical setting to estimate the causal effects of minimum ratios on hospitals. Minimum ratios led to a 58 min increase in nursing time per patient day and 9 percent increase in the wage bill per patient day in the general medical/surgical acute care unit among treated hospitals. Hospitals responded on several margins: increased use of lower-licensed and younger nurses, reduced capacity by 16 beds (14 percent), and increased bed utilization rates by 0.045 points (8 percent). Using administrative data on discharges for acute myocardial infarction (AMI), I find a significant reduction in length of stay (5 percent) and no effect on the 30-day all-cause readmission rate. The null effect on readmissions suggests that length of stay declined not because hospitals were discharging AMI patients "quicker and sicker", rather, AMI patients recovered more quickly due to an improvement in care quality per day. [ABSTRACT FROM AUTHOR]

With many countries planning to significantly increase grid renewable energy penetration levels, we consider the role of wave energy in supply–demand matching. We investigate how incorporating wave power into an offshore wind farm affects farm power predictability, smoothness, required energy storage capacity, and cost. In this paper, we do a first-order cost analysis of an offshore farm comprised of floating wind turbines and wave energy converters that are both standalone and combined and onshore compressed air energy storage. Then, we do a parameter sweep investigation of an isolated power network supplied by varied grid renewable energy penetration levels supplemented by natural gas, varied distribution of renewable energy between wind and wave power, and varied power capacity of a compressed air energy storage system supplying power to a shoreline community. For each parameter set, we consider the historical hourly electricity demand and wind-sea data of a coastal California community over a year, and optimize the energy storage schedule to reduce curtailed power, stored energy, and base gas plant operational cost. We show that a co-located wind-wave farm has smoother power supply, less energy curtailment, and higher farm-to-grid efficiency than a solely wind farm. That is, a 50%–50% wind-wave farm has a 15% smaller coefficient of variation in the power supply, 6% less curtailed power, and 2% higher farm-grid efficiency than a 100% wind farm when the grid is 100% renewable energy. These benefits of wave power potentially decrease the need for interconnecting regional transmission lines to match power supply with demand. The intent of this paper is to provide baseline system technical results to help future researchers and policy makers make decisions about offshore hybrid wind-wave-storage farms. • We do a joint parameter sweep and energy storage schedule optimization on the farm. • A co-located wind-wave farm has 15% smoother power than a wind farm. • A co-located wind-wave farm has 6% less energy curtailment than a wind farm. • A co-located wind-wave farm has 2% higher power efficiency than a wind farm. • Attaching a wave energy converter (WEC) to a wind turbine reduces WEC cost by 43%. [ABSTRACT FROM AUTHOR]

• The models are applicable to various engineer fields. • The relationships between the UCS and other mechanical indices. • Generalized function for capturing the effect of curing time. Given the wide application of lime stabilized loess in road base and subbase, subgrade, building foundations and slopes, a set of generalized and accurate mechanical estimation models of lime stabilized loess is essential for quality control. In this research, three modified models were proposed to capture the effects of lime content and compactness on the unconfined compressive strength (UCS), California bearing ratio (CBR), resilience modulus (M R), internal friction angle (φ), cohesion (c) and compression coefficient (a 1-2) of lime stabilized loess at both 7-day and 28-day curing periods. Compared with the logarithmic model, the hyperbolic model can more reasonably characterize the effect of the curing period on these mechanical indices. Taking those factors together, a set of three-factor estimation models based on both 7-day and 28-day references were developed and show high reliability. Moreover, the establishment of relationships between UCS and CBR , M R , φ and c provides a shortcut for the mechanical property estimation of lime stabilized loess. Finally, based on multiple sets of UCS data sourced from numerous papers, a hyperbolic model with a wider application range was established to link normalized UCS to curing time. The proposed estimation models can be adopted for design checking and quality evaluation of lime stabilized loess applied in various engineering fields. [ABSTRACT FROM AUTHOR]

• The use of Solution Road RomixSoilfix in granular base material is evaluated. • Deformation performance of SRX-GBM in different stress states is analyzed. • Field application and evaluation of SRX-GBM are performed in highway. Granular base material (GBM) is widely used as a base course in flexible pavement, since it has good performances to disperse stress and resist deformation. However, due to the increasing cost of pavement construction resulted by the low stiffness of GBM, the application of flexible pavement was rarely reported in Chinese expressways. This paper aimed to investigate the effect of Solution Road RomixSoilfix (SRX) on deformation performance of GBM and to evaluate the feasibility of SRX modified GBM (SRX-GBM) as a flexible base in expressway structure. A series of experimental tests were performed to analyze the influence of SRX content on curing process, stability and deformation performance of SRX-GBM. Subsequently, a flexible pavement was constructed using the SRX-GBM base course at the field sections on Tengchong to Longchuan Expressway in Southwest China. Results show that the use of SRX improves the California bearing ratio (CBR) of GBM, which raises from 145% to 417%. The dry shrinkage resistance and water stability of SRX-GBM are obviously better than cement-stabilized macadam. The results of triaxial tests demonstrate that the resilient behavior of SRX-GBM is dependent on the stress characteristics and curing times. The existing stress-related model is available for evaluating the changes of stress contribution to the Mr with the increase of SRX content. Since the base course has a high residual moisture content in a humid climate, the CBR of 250% was proposed by field CBR and deflection tests as a criterion for controlling curing time. [ABSTRACT FROM AUTHOR]

• Numerically represented a diagonal SMA-wire based bracing system. • Developed probabilistic bridge models for various uncertainties. • Proposed refined definition for failure for SMA-retrofitted bridges. • Exemplified risk-based design of SMA wire-based brace for bridges in California. In this paper, a risk-based seismic design method for diagonal self-centering braces (SCB) with shape memory alloy (SMA) wires implemented in multi-column bents is proposed to enhance the performance of older concrete bridge classes. An existing SMA-SCB system that improves the performance of older reinforced concrete building frames was adopted to numerically examine further applications to nonductile bridges. To this end, risk-based seismic design of the SMA-SCB to minimize the probability of collapse and demolition of bridge classes was developed by adopting a sophisticated finite element model that reflects the recentering behavior of an actual SMA-SCB; probabilistic bridge models considering uncertainties associated with the material and geometrical properties of bridges; refined failure condition; and the total probability theorem. A comparison of the results of the traditional seismic fragility-based and proposed designs indicated that the amount of SMA-SCB required by the former is nearly half of that required by the proposed approach because the traditional approach overemphasizes the demand at larger ground motion intensities. [ABSTRACT FROM AUTHOR]

• Assess the seismic reliability analysis of a bridge network using BC approach. • Propose a new computationally efficient BC concept to rank importance order of bridges. • Suggest the seismic retrofit plan of a bridge network based on the BC importance of bridges. • Present a BC-based approach that improves the network performance using new constructions. Although bridges are integral parts of transportation networks, they are vulnerable to natural hazards. The damage incurred during these natural hazards events leads to network disruptions and hinders post-event recovery. Therefore, improving the seismic performance of transportation networks is critical. This paper proposes an origin–destination pair betweenness centrality (BC)-based methodology to rank bridges according to their importance in a complex bridge network. The proposed methodology is an easy-to-use approach that considers user equilibrium and topology and significantly shorten the computational time. In addition, this study proposes two strategies including the retrofit of bridges and construction of new bridges based on the BC importance of bridges to mitigate the seismic risk of bride networks. A well-studied bridge network in California was used to demonstrate the application and the effect of different strategies was compared in this study. Bridges located at the shortest path significantly affect the seismic performance of bridge networks, indicating there is no need to rehabilitate or replace all bridges under budget constraints. Also, constructing five new bridges based on BC importance improved seismic performance of the bridge network much more than retrofitting 21 bridges. Moreover, the proposed methodology presents a significant reduction in computational time (only 1/30 compared with an existing study) in resolving the importance of bridges in a network and can be applied to provide useful information for disaster management teams to mitigate the seismic risk of bridge networks. [ABSTRACT FROM AUTHOR]

Globally, decentralized energy systems are gaining popularity due to their potential for energy accessibility, energy resilience, and sustainability benefits. Existing research on an energy system decentralization approach, community choice aggregation (CCA), shows its ability to lower energy costs and increase renewable electricity consumption for U.S. communities. Nevertheless, research on the relationship between CCA and distributed electricity generation development is lacking. This paper fills this gap by investigating if the CCA approach associates with distributed generation capacity interconnection in California municipalities. The finding shows that although the average capacity has increased for all municipalities throughout the study period, contrary to proponents' arguments, the CCA approach has insignificantly decreased the capacity interconnected for municipalities. It is unclear if the result is due to a lack of higher-level support for the full CCA implementation or substitution by community-owned distributed generation. Future research is necessary to determine the CCA effect comprehensively in California. With this understanding, the research could be expanded to explore how community energy approaches work towards distributed generation across the U.S. and the globe. • Statistical evaluation of the CCA-distributed generation relationship. • The finding is contrary to CCA proponents' expectations. • Results show that CCA does not increase residential distributed generation capacity. • Global, national, and state legislative and financial supports are necessary. • Need community- and household-level distributed generation tradeoff research. [ABSTRACT FROM AUTHOR]

• A lane-level VSL (LVSL) approach with actor-critic framework is proposed to provide differential VSL control. • A hybrid reward function is presented to adjust relationship between the reducing crash risk and improving traffic efficiency. • A twin delayed deep deterministic policy gradient (TD3)-based LVSL approach is established for the first time. • The TD3LVSL shows better performance than baseline models in terms of improving traffic safety and traffic efficiency. Variable speed limit (VSL) control plays a vital role in the emerging connected automated vehicle highway (CAVH) system, which can alleviate recurrent traffic congestion caused by capacity drop and accidents. However, the effect of the VSL on improving traffic efficiency is still controversial. It is necessary to study how to explore the potential benefits of the VSL by balancing its influence on reducing crash risk and enhancing traffic efficiency. To fill the technological gap above, this paper proposes a reinforcement learning-based lane-level VSL (LVSL) control approach to conduct refined traffic control on the mainlines. Firstly, an actor-critic framework is developed to generate and evaluate the discrete speed limits of each lane in continuous action space. To optimize traffic control performance, a hybrid reward function is developed by synchronously considering traffic safety and traffic efficiency of the bottleneck area. Then, to solve the overestimation bias problem of the actor-critic methods caused by function approximation error, a twin delayed deep deterministic policy gradient (TD3) method is introduced to train the framework of the LVSL method. Finally, a real-world recurrent bottleneck of the State Route 91 highway in California is simulated with consideration of the connected automated vehicles to examine the performance of the TD3-based LVSL (TD3LVSL) controller. The simulation results reveal that the proposed method is capable of reducing crash risk and improving traffic efficiency synchronously. Compared with the LVSL controller based on the deep deterministic policy gradient approach, the TD3LVSL controller shows better performance in terms of traffic safety and efficiency. These findings indicate that the proposed controller could contribute to future traffic control in a CAVH environment. [ABSTRACT FROM AUTHOR]

Ever since California experienced its energy crisis two decades ago, dynamic pricing of electricity has been the topic of discussion in numerous conferences and papers. Scores of pilots involving some 400 treatments of time-varying rates have been done to assess customer response to dynamic pricing around the globe, beginning with California's Statewide Pricing Pilot that ran from 2003 to 04. In the past few years, we are witnessing large scale deployment of static time-varying rates in states such as California and Michigan. Colorado is about to embark on that journey. However, other than OG&E's deployment of dynamic pricing, we have not seen much deployment of dynamic pricing in North America. California's power outages in August 2002 have rekindled interest in the topic. But almost all of the discussion about dynamic pricing has focused on summer peaking utilities. In this paper, we discuss the experience of Hydro Quebec, a winter peaking utility in Canada. Hydro Quebec has tested both critical-peak pricing and peak-time rebates. The results are very encouraging and quite consistent with results from summer peaking utilities. The article is based on a question and answer format with Frederic Pelletier, who advises Hydro Quebec on tariff strategy. He had posted some results from their dynamic pricing deployment on LinkedIn in response to some results I had posted from three-pilots with TOU rates that had been carried out in Maryland. I put a few questions to Frederic and this article evolved out of that conversation. In the article that follows, the questions are mine and the answers are exclusively his. [ABSTRACT FROM AUTHOR]

• • Compressed air storage alone can firm up wind and solar variability in California. • • Increased penetration of renewable energy can decrease CAES storage requirements. • • Type of renewable energy strongly affects the required CAES capacity. • • Bulk CAES deployment has limited applications due to real-world constraints. Bulk energy storage can play an important role in the decarbonization of renewable-dominant electricity system. It can offer a solution to the grid balancing problem caused by the variability in the output of renewable power generation. This paper explores the requirement for compressed air energy storage (CAES) capacity as the penetration of renewable energy increases to compensate for the variability of wind and solar. A case study for California using parsimonious macro energy model with real-world historical demand and hourly weather data has been utilized to do this analysis. In the least-cost model, with no excess wind and solar power generation, required CAES capacity is 3.71TWh in 100% decarbonized scenario. If a strict rule of net-zero curtailment was in place, the storage capacity required would be 3.2% higher (3.83TWh) with 9.8% increased cost of electricity. However, the required CAES capacity decreases with the excess solar and wind power generation. In case of 2 times increase in the wind and solar potential the required CAES capacity with strict zero curtailment would be 19.2% less (3.10TWh) and the corresponding cost would decrease up to 29.7%. The study also revealed that the type of renewable energy mix (wind, solar or both) has strong effect on the required energy storage capacity for deep decarbonization. The presented results demonstrate that excess wind and solar power generation can be used to significantly reduce the required storage needs for a fully renewable power system at reduced cost. In California, building of a large battery energy storage (up to 3 TWh) is limited by societal, geographical, and economic constraints. This study suggests that in addition to the battery energy storage California may need to look for other dispatchable power sources (or the equivalent in load flexibility) for 100% wind and solar based fully decarbonized power system. [Display omitted] [ABSTRACT FROM AUTHOR]

• Machine learning can predict the in-situ hydrocarbon viscosity at a specific depth. • High-viscosity and low-viscosity samples can be differentiated at F1 Score of 0.95. • Viscosity is predicted at a mean absolute error of 27 ∓ 3 cP. • Raw white-light and filtered UV-light images are strongly associated with viscosity. • Variances of white-light RGB channels are strongly associated with viscosity. The paper implements image processing followed by supervised learning to predict the viscosity of hydrocarbon present at a specific reservoir depth in a heavy oil reservoir located in California, USA. Viscosity is predicted from white-light and UV-light images of side-wall core samples extracted from specific depths along the wellbore. The fluid viscosity primarily ranges from 5 to 600 cP at a temperature of 180 degF. The white-light and UV-light images are available for 680 side-wall core samples. Sobel, Sato, Hessian, LBP, and Multi-Otsu filters extract meaningful features from red, green, and blue pixel intensities of the white-light and UV-light images. The distribution of pixel-wise values for each filtered/raw image is further processed to derive histogram-based features. As a result, 600 features are extracted for each side-wall core sample. Thresholds based on the variance, MI score, F score, and Person's r are imposed to select the most informative features for the viscosity prediction. Compared to filtered white-light images, raw RGB white-light images have stronger viscosity association. On the contrary, filtered UV-light images have stronger viscosity association as compared to raw RGB UV-light images. Supervised learning is deployed using the resulting features in both regression and classification of target viscosity. For the prediction of continuous-valued viscosity (regression), random forest is the best performer with a mean absolute error of 27 ∓ 3 cP. For the viscosity classification, high-viscosity and low-viscosity samples can be very well detected at a F1 Weighted Score higher than 0.95 and Matthew's Correlation Score higher than 0.9. Overall, for viscosity prediction, classification is much superior to the regression. The novel image-based viscosity prediction workflow will help lower the cost and improve the precision of laboratory-based viscosity measurements and simultaneously contribute to a high-resolution reservoir model development. [ABSTRACT FROM AUTHOR]

• Systematic review has been performed on Scopus and Web of Science (WoS) databses, and 122 studies were appraised. • Various C&D waste materials have been reviewed to find their potential usage as subbase/base material in the pavement. • Influence of C&D waste as an alternative aggregate was assessed in terms of mechanical and durability properties. • Opportunities and drawbacks associated with the application of C&D waste in the subbase/base of pavement are discussed. • Research gaps in existing literature are critically identified and future prospectives are judiciously proposed. Construction and demolition (C&D) waste consists of debris generated from construction, repair, renovation, and demolition of bridges, roads, buildings, and other infrastructure. It has certain adverse environmental impacts resulting in the consumption of valuable land space, landfill depletion, consumption of energy, and non-energy resources, resource depletion, pollution, etc. This paper presents a systematic review of literature on research pertaining to the usage of different types of C&D wastes as aggregate in the bound and unbound subbase or base layers of pavement. The literature search on two databases, namely: Web of Science (WoS) and Scopus, resulted in 7610 records. After adopting inclusion and exclusion criteria, the full text of 122 records is critically assessed. The efficacy of utilization of C&D waste in base or subbase layers in terms of mechanical and durability properties has been critically reviewed. California bearing ratio (CBR), and unconfined compressive strength (UCS) value of blended C&D waste materials are found to decrease with an increase in waste content in the mix. However, chemical stabilization could result in the enhancement of both properties. The resilient moduli for these waste materials are found to be enhanced with an increase in stabilizer dosage, curing duration, and confining pressure. Similarly, the permanent deformation of treated C&D waste materials has shown a decreasing trend with an increase in the stabilizer content. The addition of recycled asphalt pavement (RAP) aggregates to the cement-treated base reduced the flexural strength but enhanced the ductile behaviour. Research gaps are identified critically and future prospectives are proposed. Evaluation of the suitability of unexplored wastes for subbase or base layers applications is suggested. The comprehensive information enlisted in this paper may serve as a guideline for adopting non-traditional aggregates in base or subbase layer construction. Further, the outcome of this investigation is expected to be beneficial to adjudge suitable construction practices to accelerate waste utilization for sustainable pavement construction. [ABSTRACT FROM AUTHOR]

As low-cost air quality sensors become more widely utilized, more tools and methods are needed to help users access/process sensor data, identify poorly performing sensors, and analyze/visualize sensor data. Free and open-source software (FOSS) packages developed for use on FOSS data science platforms are well-suited to support this need by offering replicable and shareable tools that can be adapted to meet a user or project's specific needs. This paper describes enhancements to the FOSS AirSensor R package (version 1.0) and the DataViewer web application (version 1.0.1) that have been developed to support data access, processing, analysis, and visualization for the PurpleAir PA-II sensor. This paper also demonstrates how these enhancements may be used to track and assess the health of air sensors in real-time or for large historical datasets. The dataset used for this analysis was collected during a multi-year project (with sensors deployed from October 2017 to October 2020) involving the distribution of approximately 400 PA-II sensors across 14 communities in southern, central, and northern California. Applying the tools in the AirSensor package revealed a dramatic variability in sensor performance, mainly driven by seasonal trends or particulate matter source type. These results also indicate that this sensor can provide useful data for at least three years with little evidence of substantial or consistent drift. Further, high agreement was observed between co-located sensors deployed at different times, indicating that it may be reasonable to compare data from old and new PA-II sensors. In addition to assessing the long-term performance and reliability of the PA-II sensor, this analysis serves as a model for how data from large sensor networks may be effectively processed, evaluated, interpreted, and communicated. • An open-source R package has been enhanced to better support sensor data analysis. • Package is used to assess multiple years of data from multiple sensor networks. • Analysis reveals seasonal variability in the performance of air quality sensors. • Though there is little indication of substantial or consistent drift over time. • Example illustrates how the package can be used with real-time or historical data. [ABSTRACT FROM AUTHOR]

Applying synthetic fibers is one of the most effective physical methods for soil improvement and stabilization. In this paper, we determine the effects of polypropylene (PP) and poly vinyl alcohol (PVA) fibers contents (0.2, 0.5, 1.0 and 1.5%) at environment temperature (25 °C) and various freezing soil temperatures (0, −5, −10 and − 15 °C) on stabilization of the clayey sand soil (SC) samples. The results show that for both types of fibers, by simultaneous increasing of fiber content and lowering the temperature, the values of the unconfined compressive strength (UCS), elastic modulus (EM), ductility and California bearing ratio (CBR) increase, and they all reach their maximum values for 1.5% fiber content at −15 °C. Moreover, we show that PVA fibers perform better compared to PP fibers in increasing of UCS, EM, ductility and CBR values for different fiber contents and temperatures. This can be attributed to the fact that PVA has higher tensile and flexural strength compared to PP fiber. • The effect of freezing soil temperatures (FSTs) on strength parameters of a clayey sand (SC) soil stabilized by poly propylene (PP) and poly vinyl alcohol (PVA) fibers is investigated. • For both types of PP and PVA fibers, by simultaneous increasing of fiber content and lowering the temperature (FSTs effect), the strength parameters of SC soil samples are increased. • The strength parameters of SC soil samples reinforced by PP and PVA fibers reach their maximum values for 1.5% fiber content at −15 °C. • The PVA fibers perform better compared to PP fibers in increasing of strength parameters for different fiber contents at environment temperature and FSTs. [ABSTRACT FROM AUTHOR]

Imperial Valley, California is called 'America's winter salad bowl' as the agricultural sector produces 2/3 of the winter vegetables eaten in the U.S. However, there is little understanding of the interplay of policy, environmental conditions, and economic components of the agricultural system producers face in the valley. Using a resilience approach, this study asks the research question of how producers in the Imperial Valley perceive and are adapting to changing conditions. Through interviews with eight agricultural producers, this paper identifies current challenges and strategies producers employ to navigate a changing agricultural system and build resilience for the future. Results highlight the critical role of government policies in the establishment of agriculture in the Valley, while producers see recent policies, such as labor and food safety regulations, as increasing their operation costs. Producers face economic challenges including losing contracts to Mexico, market price fluctuations, land leasing, and the COVID-19 pandemic. Maintaining access to water resources and investing in mechanization and technological innovations are two strategies producers use to maintain resilience and navigate change. Understanding how producers remain resilient given ongoing, and often accelerating, political, environmental, and economic change is critical to maintain agricultural systems in California and beyond. [Display omitted] • Explores producer perspectives to understand policy, environmental, and economic change within an agricultural system. • Government policy has negative and positive impacts on producer resilience, depending on the policy and context. • Labor and food safety regulations are major policy drivers of change, increasing operating costs. • Producers secure water access and invest in mechanization to build resilience to future change. • Understanding how producers navigate change is critical to maintaining agricultural systems. [ABSTRACT FROM AUTHOR]

• Propose a novel deep learning model based on Runge-Kutta recurrent neural network (RKRNN) with prior knowledge to realize structural system identification and seismic response prediction. • Formulate a partition training strategy to train the proposed neural network to improve the efficiency of training. • Utilize three numerical examples to valid the feasibility of RKRNN model including a linear three degrees of freedom system, a nonlinear single degree of freedom system with Bouc-Wen hysteresis and a simply supported bridge. • Site monitoring data from a bridge located in California has been used to further validate the proposed approach. In the seismic analysis of structural systems, dynamic response prediction is an essential problem and is significant in every stage during the structural life cycle. Conventionally, response analysis is carried out by numerical analysis. However, when the structural parameter is unknown, the establishment of a numerical model will be difficult. Enlightened by the Runge-Kutta (RK) numerical algorithm, this paper proposes a novel recurrent neural network named Runge-Kutta recurrent neural network (RKRNN) to realize the seismic response prediction. A partition training strategy is formulated to train the proposed neural network and to improve the efficiency of training. The proposed model can be trained by using a limited number of samples. Three numerical examples are utilized to validate the feasibility of RKRNN model including a linear three degrees of freedom (DOFs) system, a nonlinear single DOF system with Bouc-Wen hysteresis, and a numerical reinforced concrete bridge model. Additionally, the site monitoring data from a real-world bridge is utilized to further validate the proposed network. The results show that the proposed RKRNN model can effectively and efficiently predict the structural response under seismic load and exhibits robustness to noise, with good potential for applications in engineering practice. [ABSTRACT FROM AUTHOR]