Continental United States Research Articles

Standardized benchmark of historical compound wind and solar energy droughts across the Continental United States

As we move towards a decarbonized grid, reliance on weather-dependent energy increases as does exposure to prolonged natural resource shortages known as energy droughts. Compound energy droughts occur when two or more predominant renewable energy sources simultaneously are in drought conditions. In this study we present a methodology and dataset for examining compound wind and solar energy droughts as well as the first standardized benchmark of energy droughts across the Continental United States (CONUS) for a 2020 infrastructure. Using a recently developed dataset of simulated hourly plant level generation which includes thousands of wind and solar plants, we examine the frequency, duration, magnitude, and seasonality of energy droughts at a variety of temporal and spatial scales. Results are presented for 15 Balancing Authorities (BAs), regions of the U.S. power grid where wind and solar are must-take resources by the power grid and must be balanced. Compound wind and solar droughts are shown to have distinct spatial and temporal patterns across the CONUS. BA-level load is also included in the drought analysis to quantify events where high load is coincident with wind and solar droughts. We find that energy drought characteristics are regional and the longest droughts can last from 16 to 37 continuous hours, and up to 6 days. The longest hourly energy droughts occur in Texas while the longest daily droughts occur in California. Compound energy drought events that include load are more severe on average compared to events that involve only wind and solar. In addition, we find that compound high load events occur more often during compound wind and solar droughts that would be expected due to chance. The insights obtained from these findings and the summarized characteristics of energy drought provide valuable guidance on grid planning and storage sizing at the regional scale.

Effects of simulated secondary organic aerosol water on PM1 levels and composition over the US

Abstract. Water is a key component of atmospheric aerosol, affecting many aerosol processes including gas-to-particle partitioning of semi-volatile compounds. Water related to secondary organic aerosol (SOAW) is often neglected in atmospheric chemical transport models and is not considered in gas-to-particle partitioning calculations for inorganic species. We use a new inorganic aerosol thermodynamics model, ISORROPIA-lite, which considers the effects of SOAW, to perform chemical transport model simulations for 1 year over the continental United States to quantify its effects on aerosol mass concentration and composition. SOAW can increase average fine aerosol water levels by up to a factor of 2 when secondary organic aerosol (SOA) is a major PM1 component. This is often the case in the south-eastern US, where SOA concentrations are higher. Although the annual average impact of this added water on total dry PM1 concentrations due to increased partitioning of nitrate and ammonium is small (up to 0.1 µg m−3), total dry PM1 increases of up to 2 µg m−3 (with nitrate levels increases of up to 200 %) can occur when RH levels and PM1 concentrations are high.

Using a time-varying LCAR-based strategy to investigate the spatiotemporal pattern of association between short-term exposure to particulate matter and COVID-19 incidence: a case study in the continental USA.

Numerous studies have demonstrated that short-term exposure to particulate matter less than 10 μm (PM10) is positively associated with the COVID-19 incidence. However, no study has investigated the spatiotemporal pattern in this association, which plays important roles in identifying high-susceptibility regions and stages of epidemic. In this work, taking the 49 native states in America as an example, we used an advanced strategy to investigate this issue. First, time-series generalized additive model (GAM) were independently constructed to obtain the state-specific associations between short-term exposure to PM10 and the daily COVID-19 cases from 1 April 2020 to 31 December 2021. Then, a Leroux-prior-based conditional autoregression (LCAR) was used to spatially smoothen the associations. Third, the temporal variation of association and the reasons underlying the spatiotemporal heterogeneity were investigated by incorporating the time-varying GAM into LCAR. Results showed that PM10 was adversely associated with COVID-19 incidence in all the states. On average, a 10 μg/m3 increase of PM10 was associated with a 7.38% (95% CI 5.20-9.64%) increase in COVID-19 cases. A substantial spatial heterogeneity was observed, with strong associations in the middle and northeastern regions and weak associations in the western regions. The temporal trend of association presented a U shape, with the strongest association in the end of 2021. The vaccination rate was examined as a significant effect modifier. Our study provided the first evidence about the spatiotemporal pattern in PM10-COVID-19 associations and suggested that air pollution deserves more attention in the post-pandemic era and in the middle and northeastern regions in America for COVID-19 control and prevention.

Forest habitats and plant communities strongly predicts Megachilidae bee biodiversity.

Megachilidae is one of the United States' most diverse bee families, with 667 described species in 19 genera. Unlike other bee families, which are primarily ground nesters, most megachilid bees require biotic cavities for nesting (i.e., wood, pithy stems, etc.). For this group, the availability of woody-plant species may be as important as nectar/pollen resources in maintaining populations. We studied Megachilidae biodiversity in the continental United States. We confirmed that the highest species richness of Megachilidae was in the southwestern United States. We examined the relationship between species richness and climate, land cover, tree species richness, and flowering plant diversity. When examining environmental predictors across the conterminous United States, we found that forested habitats, but not tree diversity, strongly predicted Megachilidae richness. Additionally, Megachilidae richness was highest in areas with high temperature and low precipitation, however this was not linearly correlated and strongly positively correlated with flowering plant diversity. Our research suggests that the availability of nesting substrate (forested habitats), and not only flowering plants, is particularly important for these cavity-nesting species. Since trees and forested areas are particularly susceptible to climate change, including effects of drought, fire, and infestations, nesting substrates could become a potential limiting resource for Megachilidae populations.

Machine Learning Downscaling of SoilMERGE in the United States Southern Great Plains

SoilMERGE (SMERGE) is a root-zone soil moisture (RZSM) product that covers the entire continental United States and spans 1978 to 2019. Machine learning techniques, Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and Gradient Boost (GBoost) downscaled SMERGE to spatial resolutions straddling the field scale domain (100 to 3000 m). Study area was northern Oklahoma and southern Kansas. The coarse resolution of SMERGE (0.125 degree) limits this product’s utility. To validate downscaled results in situ data from four sources were used that included: United States Department of Energy Atmospheric Radiation Measurement (ARM) observatory, United States Climate Reference Network (USCRN), Soil Climate Analysis Network (SCAN), and Soil moisture Sensing Controller and oPtimal Estimator (SoilSCAPE). In addition, RZSM retrievals from NASA’s Airborne Microwave Observatory of Subcanopy and Surface (AirMOSS) campaign provided a nearly spatially continuous comparison. Three periods were examined: era 1 (2016 to 2019), era 2 (2012 to 2015), and era 3 (2003 to 2007). During eras 1 and 2, RF outperformed XGBoost and GBoost, whereas during era 3 no model dominated. Performance was better during eras 1 and 2 as opposed to the pre-L band era 3. Improvements across all eras, regions, and models realized from downscaling included an increase in correlation from 0.03 to 0.42 and a decrease in ubRMSE from −0.0005 to −0.0118 m3/m3. This study demonstrates the feasibility of SMERGE downscaling opening the prospect for the development of a long-term RZSM dataset at a more desirable field-scale resolution with the potential to support diverse hydrometeorological and agricultural applications.

Growing Kalo (Taro) in the Continental United States.

As an increasing number of Native Hawaiian and Pacific Islander adults move to the continental United States, the development and implementation of resources that promote access to cultural foods and support food sovereignty on the continent is crucial to perpetuate cultural practice and connection to the 'āina (land that feeds). Kalo (taro) is an important cultural food central to Native Hawaiian identity. Native Hawaiians connect their genealogy as far back to the cultivation of kalo and the creation of kalo itself. In this practice note, we describe the creation of a māla kalo (cultivated field for taro) in Oregon by the Ka'aha Lāhui O 'Olekona Hawaiian Civic Club. An ongoing project over the past 3 years, the creation of a māla kalo exceeded expectations. Not only did the māla allow the cultivation of kalo outside of Hawaii, the māla became a place for the community to unite toward common goals of connecting with the land, promoting mental health, and creating a sense of place in their diaspora. This project indicates that not only is the creation of māla kalo in Oregon feasible, it may also be an important opportunity for the growing number of Native Hawaiians and Pacific Islanders adults living on the continent to improve health outcomes through connections with cultural foods and practices.

Bias learning improves data driven models for streamflow prediction

Study regionAndun river basin of Southern China and 273 watersheds across the continental United States. Study focusOwing to the data incompleteness and the changing environment, there is bias existing in the data driven models for streamflow prediction, greatly limiting their application in actual practice. In this paper, we incorporate the bias learning components into the data driven models to establish the mapping-bias-learning models, and design two groups of experiments, one in the Andun River basin in China and the other in 273 watersheds in the continental United States (CONUS). In the first group of experiments, we respectively apply three machine learning algorithms and one traditional statistical method to generate sixteen mapping-bias-learning models ae well as four mapping-learning-alone models. We also explore the effectiveness of different bias learning strategies, including multiple bias learning, stacking bias learning and incremental updating bias learning. In the second group of experiments, we apply the mapping-bias-learning models to the streamflow prediction in 273 watersheds of CONUS to verify the universality and robustness of the bias learning method. New hydrological insights for the regionThe mapping-bias-learning models significantly outperform the mapping-learning-alone models, and the machine learning methods are superior to the traditional statistical method in term of the ability of bias learning. In addition, adopting the appropriate bias learning strategies can further improve the streamflow forecast performance of data-driven models. The bias learning method is applicable to various watersheds with different hydrological conditions and play efficient roles at different streamflow levels. The median of MRE drops from 0.156 for mapping-learning-alone models to 0.131 for mapping-bias-learning models, while median of NSE increases from 0.918 for mapping-learning-alone models to 0.932 for mapping-bias-learning models. More inspiringly, the bias learning method can still play roles in data driven models under the environmental conditions that never occurred before, suggesting its remarkable superiority. This study highlights the power of the bias learning method for data driven models in streamflow prediction, and the promising prospect of the combination between bias learning and machine learning methods in hydrological modelling.

Assessing the suitability of lakes and reservoirs for recreation using Landsat 8

Water clarity has long been used as a visual indicator of the condition of water quality. The clarity of waters is generally valued for esthetic and recreational purposes. Water clarity is often assessed using a Secchi disk attached to a measured line and lowered to a depth where it can be no longer seen. We have applied an approach which uses atmospherically corrected Landsat 8 data to estimate the water clarity in freshwater bodies by using the quasi-analytical algorithm (QAA) and Contrast Theory to predict Secchi depths for more than 270 lakes and reservoirs across the continental US. We found that incorporating Landsat 8 spectral data into methodologies created to retrieve the inherent optical properties (IOP) of coastal waters was effective at predicting in situ measures of the clarity of inland water bodies. The predicted Secchi depths were used to evaluate the recreational suitability for swimming and recreation using an assessment framework developed from public perception of water clarity. Results showed approximately 54% of the water bodies in our dataset were classified as “marginally suitable to suitable” with approximately 31% classed as “eminently suitable” and approximately 15% classed as “totally unsuitable–unsuitable”. The implications are that satellites engineered for terrestrial applications can be successfully used with traditional ocean color algorithms and methods to measure the water quality of freshwater environments. Furthermore, operational land-based satellite sensors have the temporal repeat cycles, spectral resolution, wavebands, and signal-to-noise ratios to be repurposed to monitor water quality for public use and trophic status of complex inland waters.

Survey ofResources inBehavioral Sleep Medicine Across the Department ofDefense, Defense Health Agency.

Insomnia affects approximately 40% of active duty service members and adversely affects health, readiness, and safety. The VA/DoD Clinical Practice Guideline for the management of insomnia recommends cognitive-behavioral treatment of insomnia (CBTI) or its abbreviated version (brief behavioral treatment of insomnia [BBTI]) as the first-line insomnia treatment. The goal of this study was to assess CBTI/BBTI resources at MTFs, perceived facilitators and barriers for CBTI/BBTI, and gaps in these treatments across the Defense Health Agency. Between July and October 2022, we conducted an electronic survey of CBTI/BBTI resources across Contiguous United States and the District of Columbia (CONUS) and Outside Continental United States (OCONUS) MTFs. The survey was distributed to 154 military sleep health care providers from 32 MTFs, and a link to the survey was posted on two online military sleep medicine discussion forums. Fifteen providers from 12 MTFs volunteered to complete a 30-minute qualitative interview to explore their perception of barriers and facilitators of CBTI/BBTI at their facility. Fifty-two of 154 providers (33.8%) at 20 MTFs completed the survey. A majority of providers indicated that hypnotics remain the most common treatment for insomnia at their facility. Sixty-eight percent reported that CBTI/BBTI was available at their facility and estimated that less than 50% of the patients diagnosed with insomnia receive CBTI/BBTI. The main facilitators were dedicated, trained CBTI/BBTI providers and leadership support. Referrals to the off-post civilian network and self-help apps were not perceived as significant facilitators for augmenting insomnia care capabilities. The primary barriers to offering CBTI/BBTI were under-resourced clinics to meet the high volume of patients presenting with insomnia and scheduling and workflow limitations that impede repeated treatment appointments over the period prescribed by CBTI/BBTI protocols. Four primary themes emerged from qualitative interviews: (1) CBTI/BBTI groups can scale access to insomnia care, but patient engagement and clinical outcomes are perceived as inferior to individual treatment; (2) embedding trained providers in primary or behavioral health care could accelerate access, before escalation and referral to a sleep clinic; (3) few providers have the time to adhere to traditional CBTI protocols, and appointment scheduling often does not support weekly or bi-weekly treatment visits; and (4) the absence of quality and/or continuity of care measures dampens providers' enthusiasm for using external referral resources or self-help apps. Although there is a wide recognition that CBTI/BBTI is the first-line recommended insomnia treatment, the limited scalability of treatment protocols, clinical workflow limitations, and scarcity of trained CBTI/BBTI providers limit the implementation of the VA/DoD clinical guideline. Educating and engaging health care providers and leadership about CBTI, augmenting CBTI-dedicated resources, and adapting clinical workflows were identified as specific strategies needed to meet the current insomnia care needs of service members. Developing protocols for scaling the availability of CBTI expertise at diverse points of care, upstream from the sleep clinics, could accelerate access to care. Establishing standardized quality measures and processes across points of care, including for external providers and self-help apps, would enhance providers' confidence in the quality of insomnia care offered to service members.

Restructuring and serving web-accessible streamflow data from the NOAA National Water Model historic simulations

In 2016, the National Oceanic and Atmospheric Administration deployed the first iteration of an operational National Water Model (NWM) to forecast the water cycle in the continental United States. With many versions, an hourly, multi-decadal historic simulation is made available to the public. In all released to date, the files containing simulated streamflow contain a snapshot of model conditions across the entire domain for a single timestep which makes accessing time series a technical and resource-intensive challenge. In the most recent release, extracting a complete streamflow time series for a single location requires managing 367,920 files (~16.2 TB). In this work we describe a reproducible process for restructuring a sequential set of NWM steamflow files for efficient time series access and provide restructured datasets for versions 1.2 (1993–2018), 2.0 (1993–2020), and 2.1 (1979–2022). These datasets have been made accessible via an OPeNDAP enabled THREDDS data server for public use and a brief analysis highlights the latest version of the model should not be assumed best for all locations. Lastly we describe an R package that expedites data retrieval with examples for multiple use-cases.

National liquefaction loss database and event-level fragility functions

Liquefaction can be a significant contributor to loss due to earthquakes as observed during the Canterbury earthquake sequence or the 1995 Kobe earthquake. Geospatial liquefaction models can be used to estimate liquefaction extent after an earthquake but do not estimate liquefaction damage or impact. This article presents a liquefaction loss database for the United States and event-level fragility functions (EFFs) using aggregate liquefaction hazard measures (LHMs) derived from geospatial liquefaction models. The liquefaction loss database for the United States is developed by sampling earthquakes with Magnitude > 5.0 between the years of 1964 and 2019 in the continental United States and Alaska. Within this sample of 42 earthquakes, 11 resulted in liquefaction loss. Estimates were characterized by the type of infrastructure (e.g. transportation, utilities, and buildings) and the subcategory (e.g. for buildings: residential, commercial, and public), and then loss was estimated using the 2018-equivalent US dollar amount. When possible, loss estimates were obtained directly from the literature. Within this sample of 42 earthquakes, 6 events resulted in estimated monetary losses from liquefaction damage greater than 1% of the total event loss, including one with liquefaction damage greater than 10%. Using estimates for aggregate liquefaction hazard and population exposure derived from geospatial liquefaction models as LHMs, EFFs are presented using cost-based damage state (DS) thresholds in the United States. The fragility functions also include confidence intervals representing the uncertainty in probabilities of exceeding DS thresholds. Aggregate liquefaction hazard was found to be a preferred LHM for liquefaction loss, especially when evaluating transportation and building loss. Aggregate population exposure was found to be a better LHM for utilities. In addition, a second set of EFFs is presented using an expanded international dataset and DSs which are assigned relative to overall earthquake damage rather than cost-based DSs.

A web-based analytical framework for the detection and visualization space-time clusters of COVID-19

ABSTRACT The COVID-19 pandemic has had a profound impact worldwide and continues to spread due to various mutations of the virus. Many governmental and nonprofit agencies at different levels have quickly developed COVID-19 dashboards to disseminate information on the pandemic to the public. However, most of these systems have mainly distributed “plain” information (e.g. cases, death counts, vaccination), and rarely provided insights that can be gained from spatiotemporal analyses, such as the detection of emerging clusters. The results from these analyses hold tremendous potential for health policymakers as they try to identify ways to slow down transmission. We present a web-based geographic framework to detect and visualize space-time clusters of COVID-19. Our tightly coupled framework integrates the prospective space-time scan statistics and local indicators of spatial association (LISA) with novel 2D and 3D interactive visuals in a cyber environment (http://159.223.164.41/app/). We illustrate the applicability of our approach using COVID-19 data for the continental US. Our framework is portable to other regions that may experience infectious diseases but is also flexible to handle data of different spatial and temporal granularities. This paper fits within an effort to integrate space-time analytics for the monitoring of infectious diseases in web environment, ultimately improving health surveillance systems.

Travel distance and national access to gender-affirming surgery

BackgroundGender-affirming surgery is unequally distributed across the United States due to resource allocation, state-based regulations, and the availability of trained physicians. Many individuals seeking gender-affirming surgery travel vast distances to receive care. This study aims to quantify the distances that individuals travel to receive gender-affirming surgery based on procedure type and patient home-of-record location. MethodsPatients in the Optum Clinformatics Data Mart who underwent gender-affirming surgery were identified via Current Procedural Terminology codes. Data on patient demographics, procedural care, and location of patient and provider were collected. To be included, a patient had to meet diagnostic criteria to receive gender-affirming surgery and have a recorded surgical procedure reimbursed as part of gender-affirming surgery per the Centers for Medicare and Medicaid Services guidelines. Patients residing or receiving care outside the continental United States were excluded. Distances between the ZIP Code of each patient’s home of record and the location where the gender-affirming surgery was performed were calculated via the Google Maps Distance Matrix API. Distance traveled for gender-affirming surgery by patient state and gender-affirming surgery procedure were determined. Multivariate linear regression analysis determined predictors of distance traveled for gender-affirming surgery, whereas multivariate logistic regression identified variables associated with an increased likelihood of out-of-state travel to gender-affirming surgery. ResultsAcross 86 million longitudinal patient records, the study population included 2,743 records corresponding to 1,735 patients who received gender-affirming surgery between January 2003 and June 2020. The median distance traveled for gender-affirming surgery was 191 miles (mean: 391.5), and 36.0% of patients traveled out of their state of residence. Every patient from West Virginia, Wyoming, South Dakota, Mississippi, and Delaware traveled out of state for gender-affirming surgery. Patients with homes of record in California (18.4%), Massachusetts (20.7%), and Oregon (19.0%) were the least likely to travel out-of-state for gender-affirming surgery out of states with more than 10 gender-affirming surgery encounters. The main predictors for out-of-state travel included both feminizing and masculinizing genital surgery, as well as an insurance coverage with increased provider options. Additionally, patients traveled shorter distances for gender-affirming surgery after the post-2014 Affordable Care Act expansion compared to pre-2014. ConclusionPatients receiving gender-affirming surgery in the United States travel great distances for their care, often receiving their care from out-of-state providers. Restrictive guidelines imposed by state laws on both the access to and provision of gender-affirming surgery compound the myriad of common difficulties that patients face. It is imperative to discuss potential factors that may mitigate these barriers for those who require gender-affirming surgery.

Use of the Pfizer Respiratory Syncytial Virus Vaccine During Pregnancy for the Prevention of Respiratory Syncytial Virus-Associated Lower Respiratory Tract Disease in Infants: Recommendations of the Advisory Committee on Immunization Practices - United States, 2023.

Respiratory syncytial virus (RSV) is the leading cause of hospitalization among U.S. infants. Nirsevimab (Bevfortus, Sanofi and AstraZeneca) is recommended to prevent RSV-associated lower respiratory tract infection (LRTI) in infants. In August 2023, the Food and Drug Administration (FDA) approved RSVpreF vaccine (Abrysvo, Pfizer Inc.) for pregnant persons as a single dose during 32-36 completed gestational weeks (i.e., 32 weeks and zero days' through 36 weeks and 6 days' gestation) to prevent RSV-associated lower respiratory tract disease in infants aged <6 months. Since October 2021, CDC's Advisory Committee on Immunization Practices (ACIP) RSV Vaccines Pediatric/Maternal Work Group has reviewed RSV epidemiology and evidence regarding safety, efficacy, and potential economic impact of pediatric and maternal RSV prevention products, including RSVpreF vaccine. On September 22, 2023, ACIP and CDC recommended RSVpreF vaccine using seasonal administration (i.e., during September through end of January in most of the continental United States) for pregnant persons as a one-time dose at 32-36 weeks' gestation for prevention of RSV-associated LRTI in infants aged <6 months. Either maternal RSVpreF vaccination during pregnancy or nirsevimab administration to the infant is recommended to prevent RSV-associated LRTI among infants, but both are not needed for most infants. All infants should be protected against RSV-associated LRTI through use of one of these products.

Insulin-mediated endothelin signaling is antiviral during West Nile virus infection.

Arboviruses, particularly those transmitted by mosquitoes, pose a significant threat to humans and are an increasing concern because of climate change, human activity, and expanding vector-competent populations. West Nile virus is of significant concern as the most frequent mosquito-borne disease transmitted annually within the continental United States. Here, we identify a previously uncharacterized signaling pathway that impacts West Nile virus infection, namely endothelin signaling. Additionally, we demonstrate that we can successfully translate results obtained from D. melanogaster into the more relevant human system. Our results add to the growing field of insulin-mediated antiviral immunity and identify potential biomarkers or intervention targets to better address West Nile virus infection and severe disease.

The Development and Initial Capabilities of ThunderCast, a Deep Learning Model for Thunderstorm Nowcasting in the United States

Abstract This paper presents the Thunderstorm Nowcasting Tool (ThunderCast), a 24-h, year-round model for predicting the location of convection that is likely to initiate or remain a thunderstorm in the next 0–60 min in the continental United States, adapted from existing deep learning convection applications. ThunderCast utilizes a U-Net convolutional neural network for semantic segmentation trained on 320 km × 320 km data patches with four inputs and one target dataset. The inputs are satellite bands from the Geostationary Operational Environmental Satellite-16 (GOES-16) Advanced Baseline Imager (ABI) in the visible, shortwave infrared, and longwave infrared spectra, and the target is Multi-Radar Multi-Sensor (MRMS) radar reflectivity at the −10°C isotherm in the atmosphere. On a pixel-by-pixel basis, ThunderCast has high accuracy, recall, and specificity but is subject to false-positive predictions resulting in low precision. However, the number of false positives decreases when buffering the target values with a 15 km × 15 km centered window, indicating ThunderCast’s predictions are useful within a buffered area. To demonstrate the initial prediction capabilities of ThunderCast, three case studies are presented: a mesoscale convective vortex, sea-breeze convection, and monsoonal convection in the southwestern United States. The case studies illustrate that the ThunderCast model effectively nowcasts the location of newly initiated and ongoing active convection, within the next 60 min, under a variety of geographical and meteorological conditions. Significance Statement In this research, a machine learning model is developed for short-term (0–60 min) forecasting of thunderstorms in the continental United States using geostationary satellite imagery as inputs for predicting active convection based on radar thresholds. Pending additional testing, the model may be able to provide decision-support services for thunderstorm forecasting. The case studies presented here indicate the model is able to nowcast convective initiation with 5–35 min of lead time in areas without radar coverage and anticipate future locations of storms without additional environmental context.

416EMF Emergency Department Utilization and Hospitalizations Associated With Flooding in the Continental United States: 2008-2017

416EMF Emergency Department Utilization and Hospitalizations Associated With Flooding in the Continental United States: 2008-2017

MANGO: An Optical Network to Study the Dynamics of the Earth's Upper Atmosphere

AbstractThe Mid‐latitude All‐sky‐imaging Network for Geophysical Observations (MANGO) employs a combination of two powerful optical techniques used to observe the dynamics of Earth's upper atmosphere: wide‐field imaging and high‐resolution spectral interferometry. Both techniques observe the naturally occurring airglow emissions produced in the upper atmosphere at 630.0‐ and 557.7‐nm wavelengths. Instruments are deployed to sites across the continental United States, providing the capability to make measurements spanning mid to sub‐auroral latitudes. The current instrument suite in MANGO has six all‐sky imagers (ASIs) observing the 630.0‐nm emission (integrated between ∼200 and 400 km altitude), six ASIs observing the 557.7‐nm emission (integrated between ∼90 and 100 km altitude), and four Fabry‐Perot interferometers measuring neutral winds and temperature at these wavelengths. The deployment of additional imagers is planned. The network makes unprecedented observations of the nighttime thermosphere‐ionosphere dynamics with the expanded field‐of‐view provided by the distributed network of instruments. This paper describes the network, the instruments, the data products, and first results from this effort.

Sectoral water use responses to droughts and heatwaves: analyses from local to global scales for 1990–2019

Water use for various sectors (e.g. irrigation, livestock, domestic, energy and manufacturing) is increasing due to a growing global population and economic development. Additionally, increases in frequency and severity of droughts, heatwaves and compound drought-heatwave events, also lead to responses in sectoral water use and a reduction in water availability, intensifying water scarcity. However, limited knowledge exists on the responses in sectoral water use during these hydroclimatic extremes. In this study we quantify the impacts of droughts, heatwaves and compound events on water use of irrigation, livestock, domestic, energy and manufacturing sectors at global, country and local scales. To achieve this, datasets of reported and downscaled sectoral water use (i.e. withdrawal and consumption) were evaluated during these hydroclimatic extremes and compared to normal (non-extreme) periods for 1990–2019. Our analysis shows that these hydroclimatic extremes affect water use patterns differently per sector and region. Reported data show that domestic and irrigation water use increases during heatwaves in Eastern Europe and central continental United States, while water use decreases for thermoelectric sector, particularly in Europe while it increases in north and Eastern Asia. Additionally, global water use response patterns reveal that irrigation and domestic sectors are mostly prioritized over livestock, thermoelectric and manufacturing. Reported local-scale data reveal that for most sectors and regions/locations, stronger water use responses are found for heatwaves and compound events compared to impacts during hydrological droughts. Our outcomes provide improved understanding of sectoral water use behaviour under hydroclimatic extremes. Nonetheless, given the future threats to water availability and the limited accessible information of water use, there is an urgency to collect more monitored-driven data of sectoral water use for improved assessments of water scarcity under these extremes. Consequently, this research reveals the necessity of more realistic water use models to better represent the sectoral responses to hydroclimatic extremes.

Serviceability evaluation of highway tunnels based on data mining and machine learning: A case study of continental United States

The appraisal of a tunnel service condition usually takes into account the lining deformations and defects of the lining structure, with the assessment technique heavily rely on the expertise of the evaluation. In this study, based on the US National Tunnel Inventory (NTI) database, a TOPSIS-CRITIC evaluation system was established to comprehensively evaluate the service performance of tunnels during the operation period from three categories of structural, traffic & civil, and non-structural components, resulting in the Tunnel Service Performance Index (TSPI). Then, a K-means clustering method was used to unevenly divide the TSPI of tunnels into service levels I ∼ IV. After consulting with experts, 23 important features were selected from 88 basic characteristics of tunnels, and further refined this list to 19 key factors using Spearman correlation analysis. These features were grouped into five categories: traffic, geological, space & time, planning & construction, and operational. Due to the imbalance of categories, SMOTE oversampling method was employed for data augmentation. Further, a TPE Bayesian optimized LightGBM classifier was proposed to predict the service level of tunnels and compared with nine common machine learning models (LR, KNN, SVM, DT, RF, GBDT, XGB, ET and MLP). After comprehensive experiments and numerical testing, the TPE-LGBM classifier achieved the highest overall performance among the ten models. Lastly, the SHapley Additive exPlanations (SHAP) interpretability model in conjunction with GIS spatial geographic information was employed to conduct an in-depth analysis of how full life cycle features impact tunnel service performance during operation.