A document by John Louie. Click on the document to view its contents.

Ground magnetic observatories measure the Earth’s magnetic field and its coupling with the solar wind responsible for ionospheric and magnetospheric current systems. Predicting effects of solar- and atmospheric-driven disturbances is a crucial task. Using data from the magnetic observatory Chambon-la-Forêt at mid-latitude, we investigate the capability of our developed deep artificial neural networks in the modeling of the contributions above 24 hours and the daily variations. Two neural networks were built with the long short-term memory architecture with multiple layers. Using the data from 1995 onwards, the neural networks were trained with physical parameters indicative of solar variabilities and geographical daily and seasonal variations. By excluding the secular variation owing to the change of the Earth’s intrinsic magnetic field, we demonstrate that our approach can model the observed signals with overall good agreements for both a solar-quiet period in 2009 and a solar-active period in 2012. Particularly, using the walk forward training, we updated our models with new data leading up to the test year. The implication of this work is twofold. First, our approach can be adapted for near real-time prediction of intensity of solar and atmospheric disturbances. Second, the neural networks can be used to model the quiet variations when excluding the solar variabilities with important applications in the calculation of magnetic activity indices. This work is a proof-of-concept that deep neural networks can be used to model solar- and atmospheric-driven perturbations modulated by daily and seasonal variations as recorded at a ground magnetic observatory.

We investigate the tidal dissipation in Io’s hypothetical fluid magma ocean using a new approach based on the solution of the 3D Navier-Stokes equations. Our results indicate that Io may have experienced a period of intense tidal heating (104 TW) accompanied by excessive volcanism in the equatorial region, leading to catastrophic resurfacing of the pre-existing terrain. Tidal heating in Io’s magma ocean does not correlate with the distribution of hot spots, and is maximum for an ocean thickness of about 1 km and a viscosity of less than 104 Pa s. Due to the Coriolis effect, the k2 Love number can depend on the harmonic order. We show that the analysis of k2 may not reveal the presence of a fluid magma ocean if the ocean thickness is less than 2 km. If the fluid layer is thicker than 2 km, k20 ≈ k22/2 ≈ 0.7.

Porosity and permeability plays an important role in the investigation of the flow ability of unconventional rock. Unconventional rock such as shale gas has extremely low permeability due to nano-scale pores. Non-Darcy flow typically applies to any inertial flow where the Reynolds number is higher than 1. The study of non-Darcy flow in porous rocks by experiments carried out in the laboratory is consistently characterized by high costs and time-consuming procedures. In this study, a novel method called the lattice Boltzmann method, an alternative to the laboratory method has been used for the study of non-Darcy flow of shale gas. It provides matrix permeability from pore structure considering inertial flow which causes departure from Darcy’s law. The characteristics of non-Darcy flow are significantly influenced by the pore structure of a porous medium, with a more heterogeneous structure such as in the case of shale exhibiting a more rapid termination of Darcy flow. Two samples of shale as large as 5 mm of southern Karanpura were taken for this study. Computed micro-tomography images were acquired at 0.4 μm and 0.8 μm. The study of non-Darcy fluid flow in shale gas showed that as velocity increases, an inertial effect gets dominated in flow which results in a lowering of permeability. Earlier onset of non-Darcy behavior in complex structures is also investigated. Developing a complete understanding of the transport properties and developing an approach to assess the prospective gas flow is essential in informing the estimation of shale gas reserves and developing effective recovery strategies.

A document by Scott DeWolf. Click on the document to view its contents.

A document by Jicheng Gu . Click on the document to view its contents.

Critical Zone (CZ) scientists study the coupled chemical, biological, physical, and geological processes operating across scales to support life at the Earth's surface. In 2020, the U.S. National Science Foundation funded a network of Thematic Cluster projects called “CZ Net” to work collaboratively in answering scientific questions related to effects of urbanization on CZ processes; CZ function in semi-arid landscapes and the role of dust in sustaining these ecosystems; deep bedrock processes and their relationship to CZ evolution; CZ recovery from disturbances such as fire and flooding; and changes in the coastal CZ related to rising sea level. Data collected by these projects are diverse, ranging from time series from in situ sensors to laboratory analysis of physical samples, geophysical measurements, and others. Thus, coordinating data collection, archival, discovery, and access for the network presents significant challenges. Given the diversity in scientific domains represented, data produced, and collaborations, no single repository fully meets the needs of CZ scientists, posing questions of which repositories to use, how to enable discovery of and access to data across different repositories, and how to develop and promote best practices for sharing research products. This presentation describes cyberinfrastructure (CI) development by the CZ Net Coordinating Hub that leverages existing, domain-specific repositories for managing, curating, disseminating, and preserving data and research products from the CZ Net projects. We have developed CI that links existing data facilities and services, including HydroShare, EarthChem, Zenodo, and other repositories via a CZ Hub that provides tools for data submission, resource registry, metadata cataloging, resource discovery/access, and links to computational resources for analysis and visualization. The CZ Hub’s goal is to make data, samples, software, and other research products created by CZ Net projects Findable, Accessible, Interoperable, and Reusable (FAIR), using existing domain-specific repositories. The repository interoperability we have demonstrated for delivering data services for an interdisciplinary science program may provide a template for future development of integrated, interdisciplinary data services.

A document by Paul Rebischung. Click on the document to view its contents.

Seawater intrusion poses a substantial threat to water security in coastal regions, where numerical models play a pivotal role in supporting groundwater management and protection. However, the inherent heterogeneity of coastal aquifers introduces significant uncertainties into model predictions, potentially diminishing their effectiveness in management decisions. Data assimilation (DA) offers a solution by incorporating various types of observational data to characterize these heterogeneous coastal aquifers. Traditional DA techniques, like ensemble smoother using the Kalman formula (ESK) and Markov chain Monte Carlo, face challenges when confronted with the non-linearity, non-Gaussianity, and high-dimensionality issues commonly encountered in aquifer characterization. In this study, we introduce a novel DA approach rooted in deep learning (DL), referred to as ESDL, aimed at effectively characterizing coastal aquifers with varying levels of heterogeneity. We systematically investigate a range of factors that impact the performance of ESDL, including the number and types of observations, the degree of aquifer heterogeneity, the structure and training options of the DL models, etc. Our findings reveal that ESDL excels in characterizing heterogeneous aquifers, particularly when faced with non-Gaussian conditions. Comparison between ESDL and ESK under different experimentation settings underscores the robustness of ESDL. Conversely, in certain scenarios, ESK displays noticeable biases in the characterizing results, especially when measurement data from nonlinear and discontinuous processes are used. To optimize the efficacy of ESDL, meticulous attention must be given to the design of the DL model and the selection of training options, which are crucial to ensure the universal applicability of this DA method.

The short period fluctuations of the geomagnetic field observed in the secular acceleration have a strong presence near the equatorial region of the Earth. Origin of such fluctuations is believed to be the equatorially confined waves below the core-mantle boundary in the fluid outer core. The state of the outer core, being in a balance between the magnetic, Coriolis and buoyancy forces, can lead to the generation of such waves if a stably stratified layer exists in the outermost regions of the core. In this study, a shallow water magnetohydrodynamical model is used to investigate the characteristics of such equatorially trapped waves with a multilayered stratification and background magnetic field. A two-layer model is implemented to study the effects of radially varying background magnetic fields on the equatorially confined waves. Analytical derivation of the dispersion relation and numerical solutions have been performed to characterise these waves.The role of change in buoyancy frequency across the two layers in modifying the equatorially trapped waves is investigated.

The Advanced Ionospheric Data Assimilation (AIDA) is a real-time data assimilation model of global 3D ionosphere and plasmasphere electron density. Changes in the local space environment can occur on very short timescales, particularly during disturbed geomagnetic conditions. This space weather has an impact on many modern systems including Global Navigation Satellite System (GNSS) signals and High Frequency radio communications. To provide an ionospheric specification in real-time, AIDA ingests data streams from over 2000 GNSS receivers, using observations from both the Global Positioning System (GPS) and Galileo constellations, along with ionosonde-derived characteristics from the Global Ionosphere Radio Observatory (GIRO). These measurements are assimilated using a particle filter into the empirical NeQuick ionosphere model and Neustrelitz Plasmasphere Model (NPSM). The GNSS receiver Differential Code Biases (DCBs) are solved self-consistently using Rao-Blackwellized particle filtering. AIDA produces output at three latencies, the real-time Ultra Rapid product, the near-real-time Rapid product which operates at a 90-minute delay, and the Final product with a one day lag. The Ultra Rapid and Rapid products also include forecast products out to 6 hours ahead of real time. 

In this paper, we present numerical modeling aimed to explain Deep Long Period (DLP) events occurring in middle-to-lower crust beneath volcanoes and often observed in association with volcanic eruptions or their precursors. We consider a DLP generating mechanism caused by the rapid growth of gas bubbles in response to the slow decompression of H\textsubscript{2}O–CO\textsubscript{2} over-saturated magma. The nucleation and rapid growth of gas bubbles lead to rapid pressure change in the magma and elastic rebound of the host rocks, radiating seismic waves recorded as DLP events. The magma and host rocks are modeled as Maxwell bodies with different relaxation times and elastic moduli. Simulations of a single sill-shaped intrusion with different parameters demonstrate that realistic amplitudes and frequencies of P and S seismic waves can be obtained when considering intrusions with linear sizes of the order of 100 m. We then consider a case of two closely located sills and model their interaction. We speculate on conditions that can result in consecutive triggering of the bubble growth in multiple closely located batches of magma, leading to the generation of earthquake swarms or seismic tremors.

Estimating the radiated energy of small-to-moderate (Mw < 5) events remains challenging because their waveforms are strongly distorted during wave propagation. Even when near-source records are available, seismic waves pass through the shallow crust with strong attenuation; consequently, high-frequency energy may be significantly dissipated. Here, we evaluated the degree of energy dissipation in the shallow crust by estimating the depth-dependent attenuation (Q-1) by modeling near-source (< 12 km) waveform data in northern Ibaraki Prefecture, Japan. High-quality waveforms recorded by a downhole sensor confined by granite with high seismic velocity helped to investigate this issue. We first estimated the moment tensors for M1–4 events and computed their synthetic waveforms, assuming a tentative one-dimensional -model. We then modified the -model in the 5–20 Hz range such that the frequency components of the synthetic and observed waveforms of small events (Mw < 1.7) matched. The results show that the Q-value is 55 at depths of < 4 km and shows no obvious frequency dependence. Using the derived -model, we estimated the moment-scaled energy (eR) of 3,884 events with Mw 2.0–4.5. The median eR is 3.6×10-5 , similar to the values reported for Mw >6 events, with no obvious Mw dependence. If we use an empirically derived Q-model (~350), the median eR becomes a one-order underestimation (3.1×10-6). These results indicate the importance of accurately assuming the Q-value in the shallow crust for energy estimation of small events, even when near-source high-quality waveforms are available.

Lakes are the most prevalent and predominant water repositories on land surface. A primary objective of the Surface Water and Ocean Topography (SWOT) satellite mission is to monitor the surface water elevation, area, and storage change in Earth’s lakes. To meet this objective, prior information of global lakes, such as locations and benchmark extents, is required to organize SWOT’s KaRIn observations over time for computing lake storage variation. Here, we present the SWOT mission Prior Lake Database (PLD) to fulfill this requirement. This paper emphasizes the development of the “operational PLD”, which consists of (1) a high-resolution mask of ~6 million lakes and reservoirs with a minimum area of 1 ha, and (2) multiple operational auxiliaries to assist the lake mask in generating SWOT’s standard vector lake products. We built the prior lake mask by harmonizing the UCLA Circa-2015 Global Lake Dataset and several state-of-the-art reservoir databases. Operational auxiliaries were produced from multi-theme geospatial data to provide information necessary to embody the PLD function, including lake catchments and influence areas, ice phenology, relationship with SWOT-visible rivers, and spatiotemporal coverage by SWOT overpasses. Globally, over three quarters of the prior lakes are smaller than 10 ha. Nearly 96% of the lakes, constituting over half of the global lake area, are fully observed at least once per orbit cycle. The PLD will be recursively improved during the mission period and serves as a critical framework for organizing, processing, and interpreting SWOT observations over lacustrine environments with fundamental significance to lake system science.

In prior decades the concept of using mathematical methods to predict earthquakes was considered infeasible. Recent advances in machine learning and predictive modeling offer promising avenues to potentially realize earthquake prediction. In order to test the viability of machine learning methods, experiments were made with earthquake datasets from Kansas and Puerto Rico. The two datasets were chosen for the distinct differences in their tectonic settings. Kansas has few major faults, with a largely inactive subsurface, this produced a smaller dataset with a few large clusters. Puerto Rico is complexly faulted, with an extremely active tectonic setting, this produced a larger dataset with a large number of small clusters. In order to test the effectiveness of these two datasets for machine learning and prediction they were run through three different machine learning algorithms including an LSTM model, Bi-LSTM model, Bi-LSTM model with attention. Not only were the three different machine learning methods compared against each other for accuracy but also the datasets as well. Conclusive findings show that the two different data sets favor different processing methods. The Kansas data performs the best with the Bi-LSTM with attention model, while the Puerto Rico data performs the best with the LSTM model. This is likely due to the tectonic settings of the two regions, since the Kansas dataset has less overall data, and earthquakes are concentrated in a few large clusters, while the Puerto Rico data set has a more even distribution.

Dense and broad-coverage ocean-bottom observation networks enable us to obtain near-fault displacement records associated with an offshore earthquake. However, simple integration of ocean-bottom strong-motion acceleration records leads to physically unrealistic displacement records. Here we propose a new method using a Kalman filter to estimate coseismic displacement waveforms using the collocated ocean-bottom seismometers and pressure gauges. First, we evaluate our method using synthetic records and then apply it to an offshore Mw 6.0 event that generated a small tsunami. In both the synthetic and real cases, our method successfully estimates reasonable displacement waveforms. Additionally, we show that the computed waveforms improve the results of the finite fault modeling process. In other words, the proposed method will be useful for estimating the details of the rupture mechanism of offshore earthquakes as a complement to onshore observations.

A document by William K. Peterson. Click on the document to view its contents.

A document by Hilary Chang. Click on the document to view its contents.

Induced earthquakes are still highly unpredictable, and often caused by variations in pore fluid pressure. Monitoring and understanding the mechanisms of fluid-induced fault slip is essential for seismic risk mitigation and seismicity forecasting. Fluid-induced slip experiments were performed on critically stressed faulted sandstone samples, and the evolution of the actively sent ultrasonic waves throughout the experiment was measured. Two different fault types were used: smooth saw-cut fault samples at a 35º angle, and a rough fault created by in-situ faulting of the samples. Variations in the seismic slip velocity and friction along the fault plane were identified by the coda of the ultrasonic waves. Additionally, ultrasonic amplitudes show precursory signals to laboratory fault reactivation. Our results show that small and local variations in stress before fault failure can be inferred using coda wave interferometry for time-lapse monitoring, as coda waves are more sensitive to small perturbations in a medium than direct waves. Hence, these signals can be used as precursors to laboratory fault slip and to give insight into reactivation mechanisms. Our results show that time-lapse monitoring of coda waves can be used to monitor local stress changes associated with fault reactivation in this laboratory setting of fluid-induced fault reactivation. This is a critical first step towards a method for continuous monitoring of natural fault zones, contributing to seismic risk mitigation of induced and natural earthquakes.

Ionospheric heavy ions in the distant tail of the Earth’s magnetosphere at lunar distances are observed using the ARTEMIS mission. These heavy ions are originally produced in the terrestrial ionosphere. Using the ElectroStatic Analyzers (ESA) onboard the two probes orbiting the Moon, these heavy ions are observed as cold populations with distinct energies higher than the baseline energy of protons, with the energy-per-charge values for the heavy populations highly correlated with the proton energies. We conducted a full solar cycle survey of these heavy ion observations, including the flux, location, and drift energy, as well as the correlations with the solar wind and geomagnetic indices. The likelihood of finding these heavy ions in the preferred regions of observation called “loaded” quadrants of the terrestrial magnetotail is ~90%, regardless of the z orientation of the IMF. We characterize the ratio of the heavy ion energy to the proton energy, as well as the velocity ratio of these two populations, for events from 2010 to mid-2023. This study shows that the “common velocity” assumption for the proton and heavy ion particles, as suggested in previous work through the velocity filter effect, is not necessarily valid in this case. Challenges in the identification of the mass of the heavy ions due to the ESA’s lack of ion composition discrimination are addressed. It is proposed that at the lunar distances the heavy ion population mainly consists of atomic oxygen ions (O+) with velocities ~25% more than the velocity of the co-located proton population.

Energetic electron losses by pitch-angle scattering and precipitation to the atmosphere from the radiation belts are controlled, to a great extent, by resonant wave particle interactions with whistler-mode waves. The efficacy of such precipitation is primarily controlled by wave intensity, although its relative importance, compared to other wave and plasma parameters, remains unclear. Precipitation spectra from the low-altitude, polar-orbiting ELFIN mission have previously been demonstrated to be consistent with energetic precipitation modeling derived from empirical models of field-aligned wave power across a wide-swath of local-time sectors. However, such modeling could not explain the intense, relativistic electron precipitation observed on the nightside. Therefore, this study aims to additionally consider the contributions of three modifications – wave obliquity, frequency spectrum, and local plasma density – to explain this discrepancy on the nightside. By incorporating these effects into both test particle simulations and quasi-linear diffusion modeling, we find that realistic implementations of each individual modification result in only slight changes to the electron precipitation spectrum. However, these modifications, when combined, enable more accurate modeling of ELFIN-observed spectra. In particular, a significant reduction in plasma density enables lower frequency waves, oblique, or even quasi-field aligned waves to resonate with near $\sim1$ MeV electrons closer to the equator. We demonstrate that the levels of modification required to accurately reproduce the nightside spectra of whistler-mode wave-driven relativistic electron precipitation match empirical expectations, and should therefore be included in future radiation belt modeling.

(Revised) The streamer to leader transition defines much of the physics of long sparks near atmospheric pressures. Streamer length is an important parameter in understanding lightning protection because of its link to step length and striking distance. While streamers are routinely observed in the lab, there have been only a few observations in the field. Fewer still are of natural flashes, and almost none have been observed much above sea-level.

One of the outstanding questions in lightning research is how dart leaders (also called recoil leaders or K-leaders) initiate and develop during a lightning flash. Dart leaders travel quickly (106-107 m/s) along previously ionized channels and occur intermittently in the later stage of a flash. We have recently reported some insights into dart leader initiation and development based on our BIMAP-3D observations. In this presentation we will expand on that work by combining observations and modeling to try to understand the observed dart leader behaviors. BIMAP-3D consists of two broadband interferometric mapping and polarization (BIMAP) systems that are separated by 11.5km at Los Alamos National Laboratory. Each station maps the lightning VHF sources in a 2D space, and the combination of the 2-station measurements provides a detailed 3D source map. A fast antenna is also included at each station for electric field change measurements. Our previously reported observations suggest dart leaders commonly exhibit an initial acceleration, followed by a more gradual deceleration to a stop. We also modeled the dart leader electric field change with a simple configuration of two point-charges, finding that the modeled tip charge increased in magnitude during the initial acceleration in some simple cases. We now employ a more sophisticated model to better understand the distribution of charge along the dart leader channel, and the background electric field in which the dart leader develops.Presented at the AGU 2023 Fall Meeting

A good understanding of the rupture patterns of small earthquakes is essential to understand the differences between earthquakes of different sizes. However, resolving the source complexity of small events (Mw<5) is challenging, because their seismic waveforms are distorted during propagation. In this study, we used high-quality seismic waveforms recorded by an excellent downhole sensor in Japan to directly examine the source complexities of 64 Mw3.3-5.0 short-range earthquakes (< 8 km). We found that even the waveforms of microearthquakes (Mw < 2) were simple at the sensor, indicating that the waveforms were scarcely disturbed by structural inhomogeneities. We inferred the moment rate functions from the shapes of the direct P-waves, which showed diversity in their complexity. Even conservatively estimated, 30% of the events had multiple subevents. The results suggest that methods that account for complexity, rather than those that assume a simple source pattern, are required to characterize even small earthquakes.