We present a deep-learning approach for earthquake detection using waveforms from a seismic array consisting of multiple seismographs. Although automated, deep-learning earthquake detection techniques have recently been developed at the single-station level, they have potential difficulty in reducing false detections owing to the presence of local noise inherent to each station. Here, we propose a deep-learning-based approach to efficiently analyze the waveforms observed by a seismic array, whereby we employ convolutional neural networks in conjunction with graph partitioning to group the waveforms from seismic stations within the array. We then apply the proposed method to waveform data recorded by a dense, local seismic array in the regional seismograph network around the Tokyo metropolitan area, Japan. Our method detects more than $97$ percent of the local seismicity catalogue, with less than $4$ percent false positive rate, based on an optimal threshold value of the output earthquake probability of $0.61$. A comparison with conventional deep-learning-based detectors demonstrates that our method yields fewer false detections for a given true earthquake detection rate. Furthermore, the current method exhibits the robustness to poor-quality data and/or data that are missing at several stations within the array. Synthetic tests demonstrate that the present method has the potential to detect earthquakes even when half of the normally available seismic data are missing. We apply the proposed method to analyze 1-hour-long continuous waveforms and identify new seismic events with extremely low signal-to-noise ratios that are not listed in existing catalogs. (241words)

The eruption of the Hunga Tonga-Hunga Ha’apai volcano on 15 January 2022 triggered atmospheric waves at all altitudes. The NASA Ionospheric Connection Explorer (ICON) and ESA Swarm satellites were well placed to observe its impact on the ionospheric wind dynamo. After the Lamb wave entered the dayside, Swarm A observed an eastward and then westward equatorial electrojet (EEJ) on two consecutive orbits, each with magnitudes exceeding the 99.9th percentile of typically observed values. ICON simultaneously observed the neutral wind (90-300 km altitude) at approximately the same distance from Tonga. The observed neutral winds were also extreme (>99.9th percentile at some altitudes). The covariation of EEJ and winds is consistent with recent theoretical and observational results, indicating that the westward electrojet is driven by strong westward winds in the Pedersen region (~120-150 km). These observations imply that the dynamo is a key mechanism in the ionospheric response to the Tonga disturbance.

Although Earth’s troposphere does not superrotate in the annual-mean, for most of the year – from October to May – the winds of the tropical upper troposphere are westerly. We investigate this seasonal superrotation using reanalysis data and a single-layer model for the winds of the tropical upper troposphere. The temporal and spatial structures of the tropospheric superrotation are characterized, and the relationships between the superrotation and the leading modes of tropical interannual variability are quantified. It is also shown that the strength of the superrotation has remained roughly constant over the past few decades, despite the winds of the tropical upper troposphere decelerating (becoming more easterly) in other months. The underlying dynamics of the seasonal superrotation are studied using a combination of momentum budget analysis and numerical simulations with an axisymmetric, single-layer model of the tropical upper troposphere. Momentum flux convergence by stationary eddies accelerates the superrotation, while cross-equatorial easterly momentum transport associated with the Hadley circulation decelerates the superrotation. The seasonal modulations of these two competing factors shape the superrotation. The single-layer model is able to qualitatively reproduce the seasonal progression of the winds in the tropical upper troposphere, and highlights the northward displacement of the Intertropical Convergence Zone in the annual-mean as a key factor responsible for the annual cycle of the tropical winds.

Dry deposition of aerosols from the atmosphere is an important but poorly understood and inadequately modeled process in atmospheric systems for climate and air quality. Comparisons of currently used aerosol dry deposition models to a compendia of published field measurement studies in various landscapes show very poor agreement over a wide range of particle sizes. In this study, we develop and test a new aerosol dry deposition model that is a modification of the current model in the Community Multiscale Air Quality (CMAQ) model. The new model agrees much better with measured dry deposition velocities across particle sizes. The key innovation is the addition of a second inertial impaction term for microscale obstacles such as leaf hairs, microscale ridges, and needleleaf edge effects. The most significant effect of the new model is to increase the mass dry deposition of the accumulation mode aerosols in CMAQ. Accumulation mode mass dry deposition velocities increase by almost an order of magnitude in forested areas with lesser increases for shorter vegetation. Peak PM2.5 concentrations are reduced in some forested areas by up to 40% in CMAQ simulations. Over the continuous United States, the new model reduced PM2.5 by an average of 16% for July 2018 at the Air Quality System monitoring sites. For summer 2018 simulations, bias and error of PM2.5 concentrations are significantly reduced, especially in forested areas.

Landfalling atmospheric rivers (ARs) frequently trigger heavy and sometimes prolonged precipitation, especially in regions with favored orographic enhancement. The presence and strength of ARs are often described using the integrated water vapor (IWV) and the integrated vapor transport (IVT). However, the associated precipitation is not directly correlated with these two variables. Instead, the intensity of precipitation is mainly determined by the net convergence of moisture flux and the initial degree of saturation of the air column. In this study, a simple algorithm is proposed for estimating the heavy precipitation attributable to the IVT convergence. Bearing a strong resemblance to the Kuo-Anthes parameterization scheme for cumulus convection, the proposed algorithm calculates the large-scale primary condensation rate (PCR) as a proportion of the IVT convergence, with a reduction to account for the general moistening in the atmosphere. The amount of reduction is determined by the column relative humidity (CRH), which is defined as the ratio of IWV to its saturation counterpart. Our analysis indicates that the diagnosable PCR compares well to the forecast precipitation rate given by a numerical weather prediction model. It is also shown that the PCR in an air column with CRH < 0.50 is negligibly small. The usefulness of CRH and PCR as two complements to standard AR analysis is illustrated in three case studies. The potential application of PCR to storm classification is also explored.

Earthworms play a critical role in soil ecosystems. Analyzing the spatial structure of earthworm burrows is important to understand their impact on water flow and solute transport. Existing in-situ extraction methods for earthworm burrows are time-consuming, labor-intensive and inaccurate, while CT scanning imaging is complex and expensive. The aim of this study was to quantitatively characterize structural characteristics (cross-sectional area (A), circularity (C), diameter (D), actual length (Lt), tortuosity (τ)) of anecic earthworm burrows that were open and connected at the soil surface at two sites of different tillage treatments (no-till at Lu Yuan (LY) and rotary tillage at Shang Zhuang (SZ)) by combining a new in-situ tin casting method with three-dimensional (3D) laser scanning technology. The cross-sections of anecic earthworm burrows were almost circular, and the C values were significantly negatively correlated with D and A. Statistically, there were no significant differences in the τ values (1.143 ± 0.082 vs 1.133 ± 0.108) of anecic earthworm burrows at LY and SZ, but D (6.456 ± 1.585 mm) and A (36.929 ± 21.656 mm2) of anecic earthworm burrows at LY were significantly larger than D (3.449 ± 0.531 mm) and A (9.786 ± 2.885 mm2) at SZ. Our study showed that burrow structures at two different sites differed from each other. Soil tillage methods, soil texture and soil organic matter content at the two sites could have impacted earthworm species composition, variation of earthworm size and the morphology of burrows. The method used in this research enabled us to adequately assess the spatial structure of anecic earthworm burrows in the field with a limited budget.

The influence of shield wires on Geomagnetically Induced Currents (GICs) in power systems is considered. For the most simple power network, with one single transmission line and one shield wire connecting two substations, we derive the expressions for the voltage source and the resistance of the Thévenin equivalent circuits that, in parallel with the substations grounding resistances, produce the same effects on GICs as the full circuit. Our model extends results from previous studies that considered the effect of shield wires resistances by also including the induced geoelectric field.

The mechanisms underlying decadal variability in Arctic sea ice remain actively debated. Here we show that variability in boreal biomass burning (BB) emissions strongly influences simulated Arctic sea ice on multi-decadal timescales. In particular, we find that a strong acceleration in sea ice decline in the early 21st century in the Community Earth System Model version 2 (CESM2) is related to increased variability in prescribed CMIP6 BB emissions through summertime aerosol-cloud interactions. Furthermore, we find that more than half of the reported improvement in sea ice sensitivity to CO2 emissions and global warming from CMIP5 to CMIP6 can be attributed to the increased BB variability, at least in the CESM. These results highlight a new kind of uncertainty that needs to be considered when incorporating new observational data into model forcing, while also raising questions about the role of BB emissions on the observed Arctic sea ice loss.

Trees in seasonal climates may use water originating from both winter and summer precipitation. However, the seasonal origins of water used by trees have not been systematically studied. We used stable isotopes of water to compare the seasonal origins of water found in three common tree species across 24 Swiss forest sites sampled in two different years. Water from winter precipitation was observed in trees at most sites, even at the peak of summer, although the relative representation of seasonal sources differed by species. However, the representation of winter precipitation in trees decreased with site mean annual precipitation in both years; additionally, it was generally lower in the cooler and wetter year. Together, these relationships show that precipitation amount influenced the seasonal origin water taken up by trees across both time and space. These results suggest higher turnover of the plant-available soil-water pool in wetter sites and wetter years.

Sub-grid-scale processes occurring at or near the surface of an ice sheet have a potentially large impact on local and integrated net accumulation of snow via redistribution and sublimation. Given observational complexity, they are either ignored or parameterized over large-length scales. Here, we train random forest models to predict 1-km variability in net accumulation over the Antarctic Ice Sheet using atmospheric variables and topographic characteristics as predictors. Observations of net snow accumulation from both in situ and airborne radar data provide the input observable targets needed to train the random forest models. We find that kilometer-scale processes modify local net accumulation by as much as 172% of the atmospheric model mean. The correlation in space between the predicted net accumulation variability and satellite-derived surface-height change indicates that kilometer-scale processes operate differently through time, driven largely by the seasonal anomalies in snow accumulation.

This paper presents a novel method for obtaining the probability wave of breaking Pb of deep water, dominant wind-sea waves (that is, waves made of the energy within +-30% of the peak wave frequency) derived from Gaussian wave field theory. For a given input wave spectrum we demonstrate how it is possible to derive a joint probability density function between wave phase speed (c) and horizontal orbital velocity at wave crest (u) from which a model for Pb can be obtained. A non-linear kinematic wave breaking criterion consistent with the Gaussian framework is further proposed. Our model would allow, therefore, for application of the classical wave breaking criterion (that is, wave breaking occurs if u/c > 1) in spectral wave models which, to the authors’ knowledge, has not been done to date. Our results show that the proposed theoretical model has errors in the same order of magnitude as six other historical models when assessed using three field datasets. With optimization of the proposed model’s single free parameter, it can become the best performing model for specific datasets. Although our results are promising, additional, more complete wave breaking datasets collected in the field are needed to comprehensively assess the present model, especially in regards to the dependence on phenomena such as direct wind forcing, long wave modulation and wave directionality.

A common deviation from typical subduction models occurs when thrust sheets of oceanic-crust and upper-mantle rocks are emplaced over more buoyant continental lithosphere. The archetypal example of ophiolite obduction is the Semail ophiolite in the United Arab Emirates (UAE)-Oman orogenic belt, formed and obducted onto the Arabian continental margin during the Late Cretaceous. The Strait of Hormuz syntaxis, the northern extent of the UAE-Oman mountains, marks the transition from ocean-continent convergence in the Gulf of Oman to continental collision along the Zagros Mountains. Based on new seismic data from a focused recording network, we infer crustal and mantle deformation in the northeastern corner of the Arabian plate (including the southern Zagros and the UAE-Oman mountains), using observations from anisotropic tomography and shear-wave splitting measurements. We recover a change of ~90˚ in the axis of fast-anisotropic orientations in the crust from the Zagros to the UAE-Oman mountain belt, consistent with the dominant strike of the orogenic belts. We also find evidence for localized fossil deformation in the lithospheric mantle underlying the UAE-Oman mountain range, possibly related to stress-induced tectonism triggered by underthrusting of the proto-Arabian continental margin beneath the overriding Semail ophiolite. These orientations, averaging 15˚ anticlockwise, provide the first geophysical verification of geological

Forecasting microseismic cloud shape as a proxy of stimulated rock volume is essential for the design of an energy extraction system. The microseismic cloud created during hydraulic stimulation is known empirically to extend in the maximum principal stress direction. However, this empirical relationship is often inconsistent with reported results, and the cloud growth process remains poorly understood. This study investigates microseismic cloud growth using data obtained from a hydraulic stimulation project in Basel, Switzerland, and explores its correlation with measured in-situ stress. We applied principal component analysis to a time series of microseismic distribution for macroscopic characterization of microseismic cloud growth in two- and three-dimensional space. The microseismic cloud in addition to extending in the direction of maximum principal stress expanded to the direction of intermediate principal stress too. The orientation of the least microseismic cloud growth was stable and almost identical to the minimum principal stress direction. Following the stimulation, the orientation of microseismic cloud growth was consistent with the in-situ stress direction. Further, microseismic cloud shape ratios showed good agreement when compared with in-situ stress magnitude. The permeability tensor estimated from microseismicity also presented good correlation in terms of direction and magnitude with microseismic cloud growth. We show that in-situ stress plays a dominant role by controlling the permeability of each existing fracture in the reservoir fracture system. Consequently, microseismic cloud growth can be scaled by in-situ stress if there is sufficient variation in the existing faults.

Vertisols shrink and swell with changes in soil moisture, influencing hydraulic properties. Vertisols are often in floodplains, yet the importance of flooding as a source of soil moisture remains poorly understood. We used blue dye and deuterated water as tracers to determine the role of the macropore network in matrix recharge under artificial flood durations of 3 and 31 d in large soil monoliths extracted from a forested soil. Gravimetric soil moisture content increased by 47% in the first three days, then increased only 3.5% from day 3 to 31. Post-flood moisture content was greatest in the organic-rich, top 10 cm and was lower at 10 to 75 cm where organic matter was less. Deuterium concentration revealed that soil moisture in the top 10 cm was quickly dominated by artificial flood water, but at depth remained <80% floodwater even after 31 d. Pervasive dye staining of ped surfaces in the top 4 cm indicated connectivity to flood waters but staining at depth was less and highly variable. The isotopic composition of soil water at depth continued to shift toward flood water despite no differences in dye staining between days 3 and 31. Results indicate flooding initially but incompletely recharges matrix water via macropores and suggest the importance of flooding as a source of matrix recharge in vertic floodplain soils may depend more on flood frequency than duration. Isotopic composition of matrix water in vertic soils depends on both advective and diffusional processes, with diffusion becoming more dominant as porosity decreases.

We apply three different methods based on the analysis of the multi-component seismic data to detect seismovolcanic tremors and other seismovolcanic signals, to propose an approach to classify them and to locate their sources. We use continuous seismograms recorded during one year by 21 stations at the Piton de la Fournaise volcano (La Réunion, France). The first method allows to detect seismovolcanic signals based on stability in time of the inter-components cross-correlations function. Two other methods based on the simultaneous analysis of the whole network can be used to detect seismovolcanic signals and to locate their sources. In a second approach, the seismic wavefield is analyzed by calculating the width of the network covariance matrix eigenvalue distribution. The third method consists in performing the 3D back-projection of the inter-stations crosscorrelations in order to calculate the network response function. Simultaneous analysis of the parameters measured by the three different methods can be used to classify different types of seismovolcanic tremors. Our results demonstrate that all three methods efficiently detect seismovolcanic tremors accompanying the 2010 eruptions and the preceding pre-eruptive seismic swarms. Furthermore, methods 2 and 3 based on simultaneous analysis of the whole network detect a large number of volcanic earthquakes. Our location results show that each seismovolcanic tremor is located in a distinct region of the volcano, close to the eruptive site at a shallow depth and the preceding seismic crisis is located deeper at about the sea level under the summit crater.

Rising summer temperatures are expected to increase tree mortality rates across the western United States. Here, we analyze subfossil wood samples from Colorado dating to the last interglacial to assess the response of two common conifers to a previous warm period. The trees experienced comparable growth rates and water use efficiency during the interglacial relative to modern despite evidence from model simulations of a ~30% increase in evaporative demand during the peak of the growing season. High-resolution isotopic analysis of the wood samples show an enrichment in the late season cellulosic δ18O relative to modern samples, which we find was associated with increased reliance on summer rain. The data are consistent with other proxy data and climate model simulations showing the interglacial was associated with wetter summers across the western US. We propose enhanced summer rain during this warm period compensated for drought stress imposed by higher evaporative demand.

The mid-to-late Miocene is proposed as a key interval in the transition of the Earth’s climate state towards that of the modern-day. However, it remains a poorly understood interval in the evolution of Cenozoic climate, and the sparse proxy-based climate reconstructions are associated with large uncertainties. In particular, tropical sea surface temperature (SST) estimates largely rely on the unsaturated alkenone Uk37 proxy, which fails to record temperatures higher than 29˚C, the TEX86 proxy which has challenges around its calibration, and Mg/Ca ratios of poorly preserved foraminifera. We reconstruct robust, absolute, SSTs between 13.5 Ma and 9.5 Ma from the South West Indian Ocean (paleolatitude ~5.5˚S) using Laser-Ablation (LA-) ICP-MS microanalysis of glassy planktic foraminiferal Mg/Ca. Employing this microanalytical technique, and stringent screening criteria, permits the reconstruction of paleotemperatures using foraminifera which although glassy, are contaminated by authigenic coatings. Our absolute estimates of 24-31⁰C suggest that SST in the tropical Indian Ocean was relatively constant between 13.5 and 9.5 Ma, similar to those reconstructed from the tropics using the Uk37 alkenone proxy. This finding suggests an interval of enhanced polar amplification between 10 and 7.5 Ma, immediately prior to the global late Miocene Cooling.

Flow regimes profoundly influence river organisms and ecosystem functions, but regulatory approaches often lack the scientific basis to support sustainable water allocation. In part, this reflects the challenge of understanding the ecological effects of flow variability over different temporal and spatial domains. Here, we use a process-based distributed hydrological model to simulate 23 years of natural flow regime in 100 target bioassessment sites across the Adige River network (NE Italy), and to identify typical nivo-glacial, nivo-pluvial, and pluvial reaches. We then applied spatial stream-network models (SSN) to investigate the relationships between hydrologic and macroinvertebrate metrics while accounting for network spatial autocorrelation and local habitat conditions. Macroinvertebrate metrics correlated most strongly with maximum, minimum and temporal variation in streamflow, but effects varied across flow regime types. For example: i) taxon richness appeared limited by high summer flows and high winter flows in nivo- glacial and pluvial streams, respectively; ii) invertebrate grazers increased proportionally with the annual coefficient of flow variation in nivo-glacial streams but tended to decline with flow variation in pluvial streams. SSN models revealed that most variation in macroinvertebrate metrics was accounted for by spatial autocorrelation, although local land use and water quality also affected benthic invertebrate communities, particularly at lower elevations. These findings highlight the importance of developing environmental flow management policies in ways that reflect specific hydro-ecological and land use contexts. Our analyses also illustrate the importance of spatially-explicit approaches that account for auto-correlation when quantifying flow-ecology relationships.