The Calaveras Fault (CF) branches from the San Andreas Fault (SAF) near San Benito, extending sub-parallel to the SAF for about 50 km with only 2-6 km separation and diverging northeastward. Both the SAF and CF are partially coupled, exhibit spatially variable aseismic creep and have hosted moderate to large earthquakes in recent decades. Understanding how slip partitions among the main fault strands of the SAF system and establishing their degree of coupling is crucial for seismic hazard evaluation. We perform a timeseries analysis using more than 5 years of Sentinel-1 data covering the Bay Area (May 2015-October 2020), specifically targeting the spatiotemporal variations of creep rates around the SAF-CF junction. We derive the surface creep rates from cross-fault InSAR timeseries differences along the SAF and CF including adjacent Sargent and Quien Sabe Faults. We show that the variable creep rates (0-20 mm/yr) at the SAF-CF junction are to first order controlled by the angle between the fault strike and the background stress orientation. We further examine the spatiotemporal variation of creep rates along the SAF and CF and find a multi-annual coupling increase during 2016-2018 the subparallel sections of both faults, with the CF coupling change lagging behind the SAF by 3 to 6 months. Similar temporal variations are also observed in both b-values inferred from declustered seismicity and aseismic slip rates inferred from characteristic repeating earthquakes. The high correlation of b-value and slip-rate changes may indicate that the SAF is extremely sensitive to small stress perturbations.

The accuracy of Interferometric Synthetic Aperture Radar surface deformation solutions depends on the quality of the chosen interferogram subset. We present a method to select interferogram subsets based on unwrapping errors rather than temporal baseline thresholds. Using Sentinel-1 interferograms over the Tulare Basin, California, we show that subtle phase noise can lead to up to 31.5 mm/yr line-of-sight (LOS) errors in short temporal baseline subset solutions, while decorrelation leads to a systematic underestimation of LOS rates (up to 92.3 mm/yr) in long temporal baseline subset solutions. Our new workflow better mitigates these noise sources at the same time. In the Eagle Ford region, Texas, our strategy better reconstructed up to 11.9 cm of cumulative LOS deformation between 2017 and 2021 over a ~900 km2 region. This deformation feature can be linked to the total volume of produced oil and water through a linear relationship.

Arctic cyclones are key drivers of sea ice and ocean variability. During the 2019-2020 Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition, joint observations of the coupled air-ice-ocean system were collected at multiple spatial scales. Here, we present observations of a pair of strong mid-winter cyclones that impacted the MOSAiC site as it drifted in the central Arctic pack ice, with analytic emphasis on the second cyclone. The sea ice dynamical response showed spatial structure at the scale of the evolving atmospheric wind field. Internal ice stress and the ocean stress play significant roles, resulting in timing offsets between the atmospheric forcing and the ice response and post-cyclone inertial ringing in the ice and ocean. A structured response of sea ice motion and deformation to cyclone passage is seen, and the consequent ice motion then forces the upper ocean currents through frictional drag. The strongest impacts to the sea ice and ocean from the passing cyclone occur as a result of the surface impacts of a strong atmospheric low-level jet (LLJ) behind the trailing cold front. Impacts of the cyclone are prolonged through the coupled ice-ocean inertial response. The local impacts of the approximately 120 km wide LLJ occur over a 12 hour period or less and at scales of a kilometer to a few tens of kilometers, meaning that these impacts occur at smaller spatial scales and faster time scales than many satellite observations and coupled Earth system models can resolve.

One of the most important steps in any AI/ML application is the pre-processing of the data. The objective of this step is to project the original data in a new basis, or in a new latent space, where the different features of the problem are comparable and where their distribution covers a large range of values. Using the data in its natural basis can lead to under-performing AI/ML models. While almost all papers in our domain are careful to normalize or standardize the data, it is less frequent to see the use of simple linear PCA transformations, and even less frequent the use of more complex non-linear projections in latent spaces. Here we show how our research team is using autoencoder neural networks to perform non-linear transformations of images, simulations and time-series used in heliophyisical applications. Autoencoder transformations allow to parametrize any type of data by projecting it onto a latent space of higher or lower dimension. In these latent spaces the transformed data commonly presents better statistical properties allowing improvements in the AI/ML modeling. In addition, autoencoders are also known as generative techniques, i.e. they can be used to produce “artificial” or “synthetic” data. We will present three particular examples of the use of autoencoders: 1) parametrization of solar wind observations using standard feed forward autoencoders, 2) parametrization of magnetosphere simulations using convolutional autoencoders, and 3) parametrization and generation of solar active regions using variational convolutional autoencoders. We will show how these parametrizations can then be used for AI/ML classification and forecasting. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 776262 (AIDA).

Determining paleotemperatures in terrestrial environments are much more challenging than those in the ocean because of stratigraphic inconsistencies, strong spatial and temporal variations in temperature, and a paucity of well-tested methods. Here we utilize the ganoine scales of gars from the family Lepisosteidae to calibrate a new terrestrial paleothermometer. Gars are widespread both in the modern and in the past, as they are a freshwater fish lineage that extends back into the Cretaceous (100 Ma) and have remained relatively unchanged during that time span. Gars constantly record water temperatures, whose yearly average is closely related to mean annual temperature, in their body tissues, including scales. These scales grow continuously throughout life, are >95% hydroxyapatite and thus are highly resistant to diagenetic alteration. Oxygen isotopes in both biogenic phosphates and carbonates have been used to reconstruct environments on land with varying degrees of success. Phosphate-oxygen isotopes are more resistant to post-mortem alteration as the phosphorus-oxygen bond is stronger than the carbon-oxygen bond. We investigate the application of phosphate oxygen isotopes to gar scales by collecting scales from modern individuals from a north-south transect across the United States, exploiting the latitudinal temperature gradient in mean annual temperatures, measuring δ18Ophosphate of those scales, and comparing these values to the average δ18Owater and temperature of each locality. We compare our δ18Ophosphate calibration to previously published curves. Our work demonstrates that the δ18Ophosphate values of gar scales are robust recorders of temperature and δ18Owater.

At the Earth’s magnetopause, the Kelvin-Helmholtz (KH) instability, driven by the persistent velocity shear between the magnetosheath and the magnetosphere, has been frequently observed during northward interplanetary magnetic field (IMF) periods and considered as one of the most important candidates for transporting and mixing plasmas across the magnetopause. However, how this process interacts with magnetic field fluctuations, which persistently exist near the magnetopause, has been less discussed. Here we perform a series of 2-D fully kinetic simulations of the KH instability at the magnetopause considering a power-law spectrum of initial fluctuations in the magnetic field. The simulations demonstrate that when the amplitude level of the initial fluctuations is sufficiently large, the KH instability evolves faster, leading to a more efficient plasma mixing within the vortex layer. In addition, when the spectral index of the initial fluctuations is sufficiently small, the modes whose wavelength is longer than the theoretical fastest growing mode grow dominantly. The fluctuating magnetic field also results in the formation of the well-matured turbulent spectrum with a -5/3 index within the vortex layer even in the early non-linear growth phase of the KH instability. The obtained spectral features in the simulations are in reasonable agreement with the features in KH waves events at the magnetopause observed by the Magntospheric Multiscale (MMS) mission and conjunctively by the Geotail and Cluster spacecraft. These results indicate that the magnetic field fluctuations may really contribute to enhancing the wave activities especially for longer wavelength modes and the associated mixing at the magnetopause.

The boreal forests of Alaska have been experiencing a changing fire regime which threatens human lives and vulnerable ecosystems. Given expected increases in fire activity with climate warming, insight into the controls on fire size from the time of ignition could provide guidance for decision support. Such insight may be especially useful in cases where many ignitions occur in a short time period. Here we investigated the controls and predictability of final fire size at the time of ignition. Using decision trees, we show that ignitions can be classified as leading to small, medium, or large fires with 50.4 ± 5.2% accuracy in cross-validation. This was accomplished using two variables: vapor pressure deficit (VPD) and the fraction of spruce cover near the ignition point. The model predicted that 40% of ignitions would lead to large fires, which accounted for 75% of the total burned area. Other machine learning classification algorithms, including random forests and multi-layer perceptrons, were tested but did not outperform the simpler decision tree model. Applying the model to areas with intensive human management resulted in overprediction of large fires. The overprediction is explained by (1) suppression of those fires and (2) the fact that ignitions in more human-influenced areas occurred during periods of higher VPD on average. Overall, this type of simple classification system could offer insight into optimal resource allocation, helping to maintain a historical fire regime and protect Alaskan ecosystems.

In this study, we have developed a hyper-resolution land-surface forcing dataset (temperature, pressure, humidity, wind speed, incident longwave and shortwave radiation) from coarse resolution products using a physically-based downscaling approach. These downscaling techniques rely on correlations with landscape variables, such as topography, temperature lapse rate corrections, surface roughness and land cover. A proof-of-concept has been implemented over the Oklahoma domain, where high-resolution observations are available for validation purposes. The hourly NLDAS (North America Land Data Assimilation System) forcing data at 0.125° have been downscaled to 500m resolution over the study area during 2015. Results show that correlation coefficients between the downscaled forcing dataset and ground observations are consistently higher and biases are lower than the ones between the NLDAS forcing dataset at their native resolution and ground observations. Results are therefore encouraging as they demonstrate that the 500m forcing dataset has a good agreement with the ground information and can be adopted to force the land surface model for land state estimation. The Noah-MP land surface model is then forced with both the native resolution NLDAS dataset and the downscaled one to simulate surface and root zone soil moisture. Model outputs are compared with in situ soil moisture observations and SMAP (Soil Moisture Active Passive Mission) products at different spatial resolutions. This work will result in a radical improvement over the current state-of-the-art forcing data and will move into the era of hyper-resolution land modeling.

Intra-continental dip-slip earthquakes often occur in the orogen and rift zones with complex tectonics, providing rare opportunities to illuminate the deformation and evolution of continental structures. However, due to the sparse seismological and geodetic observations, such earthquakes are less studied. Here, we report the fault geometries of two dip-slip earthquakes recently occurred in Southern Tian Shan and the Mongolia-Baikal rift zone revealed by InSAR . The 2020 Mw6.0 Jiashi earthquake occurred in the Keping-tage fold-and-thrust belt in southwest Tian Shan. This region is seismically active, yet most well-recorded earthquakes occurred south of the mountain front. The lack of large earthquakes beneath the belt thus hinders our understanding of the orogenic process to the north. Combining InSAR measurements and relocated aftershocks, we found that a fault model involving a shallow thrust fault and two deeper faults can best reconcile the surface deformation and aftershock distribution. Stress analysis suggests that slips on the shallow fault reactivated the older basement structures at depth. Our results reflect the basement-involved shortening activated by a thin-skinned thrust faulting event with surface deformation, implying a southward orogenic process of the southwest Tian Shan. The 2021 Mw6.7 Lake Hövsgöl earthquake occurred in the Mongolia-Baikal rift zone (MBRZ), which is located in the northern tip of the northern Mongolia, and is bounded by the Tibet Plateau orogenic belt and the Siberian Platform. Using the Bayesian inversion method we derived a fault plane with two slip patches, one is mainly strike slip and the other is mainly normal slip component. The correlation between the observed and predicted displacements by the single fault model is 97.46%. Coulomb stress analysis shows that the 2021 event has a triggering effect on the western segment of the Tunka Fault to the north, where no large earthquakes have occurred since the 1905 M8+ earthquake, raising the potential for seismic risk in this region.

Snowfall has important significance in water resources management and disaster prevention worldwide. Accurate prediction of both mean and extreme snowfall is challenging because of multiple controlling mechanisms at different spatial and temporal scales. By using a 65 years long in-situ snowfall observation, we evaluated seven different machine learning algorithms for predicting monthly snowfall in the Lower Peninsula of Michigan (LPM). The Bayesian Additive Regression Trees (BART) demonstrates the best fitting (R2 = 0.88) and out-of-sample prediction skills (R2 = 0.58) for the monthly mean snowfall followed by the Random Forest model. The BART also demonstrate strong predictive skills for seasonal and the extreme monthly snowfall. Both machine learning models also demonstrate signals of key physical processes controlling the snowfall including topography, local/regional environmental factors, and teleconnections. Particularly, models with the non-parametric framework can incorporate signals from multiple scales and nonlinear responses from the snowfall to environmental factors and that substantially improved the model prediction skills. The multiscale machine learning approach provides a reliable and computationally efficient alternative approach to predict/forecast weather and climate and has potential to be applied to other extreme weather prediction scenarios.

We investigated historical changes in temperature and phytoplankton in Fuxian Lake (FX) and Dianchi Lake(DC), two plateau lakes in Southwest China. We detected sterols and alkyl diols as algal biomarkers (dinoflagellates, diatoms, eustigmatophytes, cyanobacteria, and chlorophyta) in the sediment cores by using GC-MS. We characterized sedimentary organic matter (SOM) via Rock-Eval analysis to obtain TOC, S2, the hydrogen index (HI), and refractory carbon (RC), and measured nutrients such as total phosphorus (TP) and total nitrogen (TN) in the sediment cores. The results showed that the sedimentation rates measured by 210Pb dating ranged from 0.02 to 0.30 g cm-2 yr-1 and 0.07 g cm-2 yr-1 in FX and DC, respectively. In the past 100 years, total phytoplankton biomarker concentrations ranged from 2.25 µg g-1 to 59.1 µg g-1 and from 2.20 µg g-1 to 76.4 µg g-1 in FX and DC respectively. Primary productivity derived from the S2 and HI parameters showed an increasing trend in FX, while it showed a trend of first increasing and then decreasing in DC. Cyanobacteria and chlorophyta were the two dominant phylum in the two lakes. Algal growth was controlled primarily by temperature and nutrient availability. Principal component analysis (PCA) and multiple linear regression analysis showed that the major factors controlling the historical change in phytoplankton growth were TN and temperature in FX and DC.

Tomographic analysis of Pn arrivals times—the guided P-wave propagating within the lithospheric mantle—is ideal for studying the structure of the uppermost mantle. While plate-scale seismic images of Pn wave speeds are common beneath the continents, similar scale studies have not been possible within ocean basins due to the sparsity of seismic stations. The Cascadia Initiative (CI) dataset provides the first opportunity to image spatial variations in lithospheric structure across an entire oceanic plate. We measure 2,862 Pn arrivals from local earthquakes recorded by the CI array. Our dataset provides complete coverage of both the Juan de Fuca (JdF) and deforming Gorda plates. We invert the measured arrival times for 3D variations in anisotropic P-wave velocity and hypocentral parameters. Despite surficial evidence of extensive active faulting, the velocity structure of the Gorda uppermost mantle is remarkably consistent with predictions from a conductive cooling model (attached figure). Limited deformation at mantle depths is supported by seismic anisotropy measurements that show the fast-direction of P-wave propagation rotates in concert with the magnetic anomaly lineations. This rotation may be explained by local plate kinematics without internal deformation and hydration of the shallow mantle. In contrast to Gorda, the seismic velocity structure of the JdF plate does not exhibit a clear age dependence. Three pronounced mantle low-velocity zones are found along the southern edge of the JdF plate near the termini of large pseudofaults that contributed to the formation of the Blanco Transform fault. We attribute these velocity reductions to mantle alteration by seawater. We note that within the interior of the JdF plate pseudofaults do not appear as uniformly slow features in our seismic images. Beneath the central and northern JdF plate, P-wave speeds are ~7.7 km/s out to ~4-5 Myr before abruptly increasing to ~7.9 km/s. Curiously, this transition occurs near the onset of mantle downwelling inferred from teleseismic body wave tomography and attenuation suggesting that mantle flow dynamics may influence the structure of young oceanic lithosphere. Lastly, we note that our results do not suggest a relationship between the structure of the uppermost slab mantle and segmentation of the Cascadia megathrust.

Diapycnal mixing in the ocean interior is largely fueled by internal tides. Mixing schemes that represent the breaking of internal tides are now routinely included in ocean and earth system models applied to the modern and future. However, this is more rarely the case in climate simulations of deep-time intervals of the Earth, for which estimates of the energy dissipated by the tides are not always available. Here, we present and analyze two IPSL-CM5A2 earth system model simulations of the Early Eocene made under the framework of DeepMIP. One simulation includes mixing by locally dissipating internal tides, while the other does not. We show how the inclusion of tidal mixing alters the shape of the deep ocean circulation, and thereby of large-scale biogeochemical patterns, in particular dioxygen distributions. In our simulations, the absence of tidal mixing leads to a deep North Atlantic basin mostly disconnected from the global ocean circulation, which promotes the development of a basin-scale pool of oxygen-deficient waters, at the limit of complete anoxia. The absence of large-scale anoxic records in the deep ocean posterior to the Cretaceous anoxic events suggests that such an ocean state most likely did not occur at any time across the Paleogene. This highlights how crucial it is for climate models applied to the deep-time to integrate the spatial variability of tidally-driven mixing as well as the potential of using biogeochemical models to exclude aberrant dynamical model states for which direct proxies do not exist.

InSAR time-series (InSAR-TS) analysis enables us to obtain the displacement time-series by using a number of SAR images repeatedly acquired on the area. Among the factors affecting the accuracy of the InSAR-TS analysis, this study focuses on three factors that may severely affect the signal detection limit: 1) the selection of the reference point (determining the offset in each interferogram), 2) ramp-type artifact that originate from inaccuracy in the orbit data or ionospheric disturbance, and 3) altitude-correlated tropospheric noise. Fukushima et al. (2019, Earth, Planets and Space) proposed an InSAR-TS analysis method to simultaneously solve for the displacement time-series and the error terms mentioned above as well as the error in the digital elevation model. In the proposed method, the unwrapped phase in interferograms is assumed to be composed of a linear combination of the LOS displacement, offset, planar ramp, altitude-correlated phase, and error in the used digital elevation model. A set of unwrapped small-baseline interferograms is then inverted to simultaneously obtain the displacement time-series and the parameters describing the error terms under the minimum norm condition on the displacement time-series. In this study, I applied the above-mentioned method after some updates such as introduction of the temporal constraint adopted by the NSBAS algorithm (Doin et al., 2011) and data masking, on the ALOS-2 data acquired around the Arima-Takatsuki fault zone in western Honshu, Japan. Data of four different Paths (20 and 21 from descending orbit, 127 and 128 from ascending orbit) obtained between August 2014 and March 2021 were analyzed. Some of the original interferograms contained severe noise such as a phase ramp equivalent to approximately 25 cm of LOS displacements. The average velocity field obtained by applying the method captured a relative range decrease of a few mm/year on the southern side of the fault, consistent with the results obtained from Sentinel-1 data analysis. Given the fact that the Sentinel-1 dataset had much favorable conditions (much larger number of data and much smaller ionospheric noise), the consistency in the average velocity field suggests the effectiveness of the proposed approach.

Current changes in tropical South America due to atmospheric warming, deforestation, and glacier retreat impact moisture and water exchange between the Amazon basin and the Andes. Thus, a deeper understanding of past atmospheric variability is crucial for developing strategies for climate and environmental change scenarios in this region. Within this context, we investigated an 18-year firn core drilled at the Illimani to interpret its aerosol composition (trace elements and major ions) in relation to seasonal processes, particularly atmospheric circulation over the Amazon basin. The resulting 21st-century record showed reduced Cr contamination over the Altiplano in comparison to the late 20th century, which was probably related to reduced emissions from mining activities. Sulfur records suggest the influence of volcanic eruptions in 2006 (Rabaul) and 2014 (Nyamuragira-Nyiragongo). Overall, the aerosol composition was mainly modulated by precipitation variability over the Altiplano at both annual and seasonal timescales. However, Mn was enriched due to strengthened low-level jets in the Amazon basin during the dry season, especially in 2015. This was corroborated by the reanalysis data. Furthermore, Mn, Co, and Fe showed an unprecedented peak in the record during the wet season of 2014, which was consistent with the arrival of a dust plume from Africa over Amazonia. Therefore, the Mn enrichment record can be used as a new proxy for obtaining information about the South American Low-Level Jet, and, when considered together with more elements, might also indicate snow layers that were possibly loaded with aerosols from Africa.

This year marks the centennial of the American Geophysical Union advancing Earth and space science and 89 years of hydrologic science. The last 100 years have seen science and technology dancing a harmonious and progressively accelerated waltz. Hydroelectric power generation has made widespread electrification possible, while the rise of electronics and the advent of computers have enabled hydrologists to exploit increasingly complex models. Scientists and engineers have conquered space, and now satellite-based products and remotely sensed data have become indispensable inputs for hydrometeorological forecasting. Yet several important elements that have accompanied humanity’s history – nature, culture, and people – have been relegated; and it is only very recently, in the face of modern challenges, that they have attracted substantial attention. The advent of robust, cheap, and low-maintenance sensing equipment provides unprecedented opportunities for data collection, especially in a citizen science context. While citizens have been present throughout the history of scientific practice, developments in sensing technology, data processing and visualization, and the communication of ideas and results, are creating a wide range of new opportunities for public participation in scientific research. Integrating societal knowledge with hydrologic science, however, is not only a task for the 21st century. Historically, many civilizations have developed local water harvesting and management practices that cope with water stress by using ancient and nature-based knowledge. Indeed, indigenous peoples developed solutions that were inspired and supported by nature, and use, or mimic, natural processes to contribute to improved water management and to safeguard their water security. Technological development and knowledge integration also have a more fundamental impact on the way in which hydrologic knowledge advances, how it flows between different actors, how it disrupts power relations, and thus how it influences decisions and policy-making. We envisage that, in the next century, hydrologic science will benefit from co-creating knowledge that emerges from citizens, resonates with nature, and integrates ancient wisdom.

Geophysical granular flows exert basal forces that generate seismic signals, which can be used to better monitor and model these severe natural hazards. A number of empirical relations and existing models link these signals’ high-frequency components to a variety of flow properties, many of which are inaccessible by other analyses. However, the range of validity of the empirical relations remains unclear and the models lack validation, owing to the difficulty of adequately controlling and instrumenting field-scale flows. Here, we present laboratory experiments investigating the normal forces exerted on a basal plate by dense and partially dense flows of spherical glass particles. We measured the power spectra of these forces and inferred predictions for these power spectra from the models for debris flows’ seismic signals proposed by Kean et al. (2015), Lai et al. (2018), and Farin, Tsai, et al. (2019), using Hertz theory to extend Farin, Tsai, et al. (2019)’s models to higher frequencies. Comparison of our bservations to these predictions, and to predictions derived from Bachelet (2018) and Bachelet et al. (2021)’s model for granular flows’ seismic signals, shows those of Farin, Tsai, et al. (2019)’s ‘thin-flow’ model to be the most accurate, so we examine explanations for this accuracy and discuss its implications for geophysical flows’ seismic signals. We also consider the normalisation, by the mean force exerted by each flow, of the force’s mean squared fluctuations, showing that this ratio varies by four orders of magnitude over our experiments, but is determined by the bulk inertial number of the flow.

Use of Real-Time Control (RTC) technology in Rainwater Harvesting Systems (RWH) can improve performance across water supply, flood protection, and environmental flow provision. Such systems make the most of rainfall forecast information, to release water prior to storm events and thus minimise uncontrolled overflows. To date, most advanced applications have adopted 24-hr forecast information, leaving longer-term forecasts largely untested. In this study, we aimed to predict the performance of four different RTC strategies, based on different forecast lead-time and preferred objectives. RTC systems were predicted to yield comparatively slightly less harvested rainwater than conventional passive systems, but delivered superior performance in terms of flood mitigation and delivery of environmental water for streamflow restoration. More importantly, using a 7-day rainfall forecast, the longest commercially available prediction window, was shown to enhance the ability of RTC in mitigating flood risks and delivering an outflow regime that is close to the natural (reference) streamflow. Such a finding suggests that RTC combined with 7-day forecast can enhance the functionality of rainwater harvesting systems to restore and even mimick the entire natural flow regimes in receiving streams. This also opens up a new opportunity for practitioners to implement smart technology in managing urban stormwater in a range of contexts and for a range of stream health objectives.