The Shuram excursion, an extremely negative and prolonged carbonate δ13C anomaly, is recorded globally in late Ediacaran shallow-water marine sedimentary sequences and coincided with a time of increasing ocean oxygenation. Marine planktonic microorganisms with organic cell walls, known as acritarchs, were common in the Ediacaran Period. The onset of the Shuram excursion occurred during a time of rapid acritarch cell wall diversification, and the excursion’s resolution was followed by the appearance of macroscopic calcium carbonate biomineralizing metazoans in the late Ediacaran. Accordingly, I consider that the Shuram excursion may be attributed to sedimentary accumulations of material sourced from weakly calcified acritarchs, which primitively biomineralized their cell walls by precipitating 13C-enriched marine carbonate onto 13C-depleted organic carbon. Decay of this organic carbon to authigenic carbonate after burial may have produced the low δ13Ccarb values. The carbon isotope mass balance of global carbon cycle reservoirs permits this effect to produce a Shuram excursion of any duration. An initial organic fraction of ~0.6 is required in the acritarch-derived phase, in agreement with fossil evidence that the earliest biomineralized structures contained a major fraction of organic carbon. I will discuss how fossil, petrographic and geochemical observations are consistent with this hypothesis. During the excursion, a peak globally averaged organic carbon burial fraction of ~0.4 is predicted. Burial sequestration of organic carbon in organic-rich biomineralized calcium carbonate could account for the rise in oxygen associated with the Shuram excursion.

The shear flow-interchange instability is proposed as the initiating mechanism behind substorm onset. ULF waves occurring within minutes of substorm onset are observed in the magnetotail at frequencies similar to those of the auroral beads, which are a result of a near-earth magnetospheric instability initiating current disruption in the plasma sheet. Growth rates were statistically determined as a function of wavenumber by Kalmoni et al. (2015) using ASI data from a set of substorm events. The RCM-E provides growth phase-evolved runs of background fields for stability analysis of a magnetospheric wave equation for shear flow-interchange modes derived in Derr et al. (2019), from which growth rates and dispersion relations can be calculated for comparison with the statistically-determined growth rates and frequencies of the beads. In the plasma sheet, interchange and shear flow represent a competition between Kelvin-Helmholtz instability and overall interchange stability. On average, flux-entropy increases with radial distance. As the growth phase proceeds, the middle plasma sheet becomes nearly interchange stable, but flux-entropy decreases sharply at the inner edge. Destabilizing shear is weak in the middle of the sheet but quite strong in the SAPS region, earthward of the inner edge. We examine the conditions under which shear can overwhelm interchange stability to trigger instability. Instability phenomenology will be discussed in detail, including discussion of Doppler-resonance structure and a dimensionless parameter W* for characterizing stability domains. Mapping spatial properties to the ionosphere along field lines allows for comparison of instability wavelengths with those of the auroral beads. All substorms terminate in relaxation, either because higher order nonlinearities ultimately suppress growth or due to external conditions which alter the background fields to suppress nonlinear growth. If higher order amplitude expansion terms contribute negatively at some order, then nonlinear relaxation occurs, and a method for determining field saturation values is established.

The nature of deep earthquakes with depths greater than 70 km is enigmatic because brittle failure at this high-temperature and the high-pressure regime should be inhibited. Three main hypotheses have been proposed to explain what causes deep earthquakes within the subducting slabs, dehydration embrittlement, phase transformational faulting, and thermal runaway instability. However, the existing seismological constraints can’t yet definitively distinguish between these hypotheses because the fine 3D slab structures are not well constrained in terms of slab upper interface, thickness, and internal fine layering. To better image the slabs in the Western Pacific subduction zones, this study employs a full waveform inversion (FWI) that minimizes waveform shape misfit between the synthetics and the observed waveforms from a large dataset, with 142 earthquakes recorded by about 2,400 broadband stations in East Asia. A 3-D initial model that combines two previous FWI models in East Asia (i.e., FWEA18 and EARA2014) are iteratively updated by minimizing the misfit measured from both body waves (8–40 s) and surface waves (30–120 s). Compared to the previous models, the new FWI model (EARA2020) shows much stronger wave speed perturbations within the imaged slabs with respect to the ambient mantle, with maximum perturbation of 8% for Vp and 13% for Vs. Furthermore, the slab thickness derived from EARA2020 exhibits significant downdip and along-strike variations at depths greater than 100 km. The large intra-slab deep earthquakes (Mw>6.0) appear to occur where significant slab thinning happens. This observation suggests that the significant deformation (or strain accumulation) of the slab is likely the first-order factor that controls the distribution of large deep earthquakes within the slab regardless of their triggering mechanism.

Gully erosion is an accelerator of land degradation and one of the most critical agents threatening the environment in Namibia’s north-western region. Large gullies dominating alluvial valleys expand each year during the short but intense rains, leading to a reduction of arable land and grazing areas, destruction of roads, cattle paths, agricultural facilities, and houses, prompting territorial fragmentation and the geographical isolation of local communities. In contrast, gullies can also act as a linear oasis while providing several benefits to their inhabitants. This research aims to describe the mutual influences between a large gully and the local communities in a valley extended towards the south from Opuwo, inhabited by the same native Himba families for several generations. In-situ surveys show that the gully is a general concern in the area due to the insecurity and direct physical risk it poses to humans and their domestic animals. The second factor of distress is the accelerating land degradation in the valley, leading to the disappearance of grazing areas, forcing local shepherds to travel further in their transhumance. Ortho-imagery and spatial analysis show that 10% of the houses, 25% of the Kraals, and 50% of the gardens are less than 50 meters away from the gully border, and therefore they are in current or potential risk of abandonment, forcing eventual re‑settlements and migrations. Moreover, indigenous knowledge arises that the gully also offers a few advantages, like its ability to store water during the dry season. These benefits are frequently seen as a trap or an associated risk for the animals and children getting in the gully. To this end, it is noticeable that as the gully affects the communities and its livelihoods, it also acts as a driver of development for the gully through its agricultural and livestock practices. This is evident by the appearance of the gully heads on paths, ditches, and domestic animals’ routes, along with endemic overgrazing for decades. In summary, this research identified these prevalent human-nature dynamics and attempted to provide recommendations that can reverse accelerated degradation in the long term while describing the present and potential future of the Himba people inhabiting these fragile lands in Kaokoland.

The Department of Earth & Environmental Sciences at the University of Michigan formed an Unlearning Racism in GEosciences (URGE) pod composed of six graduate students, three postdocs and eight faculty in the beginning of 2021. The department’s Diversity, Equity and Inclusion (DEI) efforts have been building in the preceding years. Our first DEI committee was formed in 2017 and increased its activity since initiation, hosting DEI discussions and initiatives with participation from students, postdocs, staff and faculty. Existing DEI activities include a Fall Preview event for prospective graduate students, DEI office hours and book discussions, adding DEI resources to the public-facing Department website, student grants for DEI related activities, and hosting workshops. The formation of an URGE pod provided new, focused energy to our DEI efforts and bolstered ongoing work by creating a bigger, critical mass of people who met regularly and were focused on action. The scope of URGE, the NSF support for it, and the interactions with other institutions that came from it, helped give our pod momentum, legitimacy, and contributed to broader departmental support for the recommendations that it produced. It also helped our department identify our most critical deficits on a DEI front and concrete ways that we will respond to them, which parallel needs articulated in a recent report from our college’s anti-racism task force, a major focus of our Dean. Actions emerging from the URGE pod include, but are not limited to, hiring a Wellness and Inclusion Advocate staff member, creation of field safety training and guidance, and building a workshop series to address issues centered on creating a culture of wellness and inclusion (anti-bias training, ally training, etc). The formation of our pod coincided with and complemented the finalization of our department’s self-study as part of a decadal strategic planning process. Many recommendations related to hiring, inclusive teaching, reporting, and deliberate mentoring practices that our URGE pod discussed were incorporated into our department’s strategic plan that was finalized in July 2021. We are eager to translate these recommendations for anti-racism work into actions, building on and contributing to the momentum and resources of the URGE community.

It took less than two decades for plastics to become notable marine contaminants following their initial commercial introductions. In the ensuing five decades, global plastics production has increased 500% each year. Currently, 300 million metric tons of plastic are produced every year much of this ends up in the ocean. In the mid-1970s it was thought that most of this plastic derived from discarded trash by ocean-going vessels. Now, we know that plastic debris in the oceans is exported by nearly all the major rivers of the world, as well as from shing, aquaculture, petrochemical, and shipping industrial sources. These plastics are long-chain polymers that are extremely durable in the environment, which is in part, why they become troublesome pollution. One practical difficulty is that in addition to the massive geographic extent of the plastic debris, the distribution is patchy in 4 dimensions (i.e., length, width, depth, time). Our approach takes advantage of the unique spectral signatures of floating plastic that can be identified in freely available Sentinel-2 imagery from the European Space Agency (ESA). Recent analytical developments identified a floating plastic index (FPI; Biermann et al. 2020) that takes a ratio of red, red edge, NIR, and SWIR bands (4, 6, 8, and 11 respectively) to discriminate between floating plastic and other floating materials such as foam, wood and vegetation. Here we present a scaled approach to identify patches of floating plastic and generate coordinates to guide cleanup efforts. The initial area of interest is 100 km on a side and is passed through hierarchical loops that 1.) separate land from water; 2.) interrogate the image files for pixels with reflectance values consistent with FPI; and, 3.) generate pixel centroid coordinates for image pixels that contain suspected floating plastic pollution. To overcome the coarse spatial (20m, bands 6 and 11) and temporal (ca. 5 days) resolution, and dynamic nearshore environment, remotely piloted aerial systems (RPAS) navigate to pixel centroids to confirm plastic accumulations. Once identified and located, updated coordinates guide navigation via autopilot in autonomous sweeper drone vessels. These sweeper drones are equipped with stereo bow-mounted hyperspectral cameras that scan the proximal waters to identify precise locations of floating plastic, plot a navigation solution, and engage the drone’s sweeper apparatus to collect the plastic. The approach integrates, distributed processing of satellite imagery in the cloud, RPAS verification of plastic locations, communication with navigation software and hardware of cleanup vessels to plot navigation solutions at the 10m scale. Finally, onboard hyperspectral cameras collect and process stereo imagery to identify precise targets which are layered on the navigation software and hardware system to direct cleanup actions. The approach is comprised of five hierarchically integrated modules that include: Coarse Location Map

People spend most of their time indoors, and therefore exposure to aerosols and precursor-gasses in the indoor environment is of extreme concern with respect to people’s health, well-being, working efficiency, and overall life quality. Such sources include both those emitted inside as well as those which may intrude from the outside. To better quantify and understand these sources and their impacts we employ multiple air quality monitors and simple models, both within and outside of various residential environments located in different cities of different development levels in China. To enhance the livability of the indoor environment, we further work to quantify two newly relevant factors: portable indoor aerosol filters and the idea of “increased ventilation”. The combined goal of this work is to understand the combination of factors leading to an improvement in Indoor Air Quality (IAQ), including but not limited to: mass transfer to/from the indoor environment, removal and/or enhancement due to the filters and ventilation, removal from the room itself, emissions, and other possible non-linearities not accounted for on this list. Concentrations of aerosols across different sizes from 0.3um to 10um are measured and analyzed at 2-minute intervals over a minimum of 20 days, allowing for an analysis that encompasses all of the possibilities of natural and anthropogenic variability typically encountered in a real environment. This includes analysis both with and without filters, under extreme outdoor loading conditions, with intense indoor emissions sources, pseudo-equilibrium conditions, and measurements made during different meteorological events, both with and without high rates of indoor to outdoor air exchange. Results show that the impacts are relatively large under both equilibrium and high event conditions, include both the magnitude of the drop as well as the time taken to achieve the reduction. A further conclusion is that increased ventilation may lead to a worsening of IAQ. A final conclusion is that larger particles more likely associated with PM2.5 are removed at a different rate from much smaller particles more associated with PM0.5 or PM1.

Multi-decadal expansion in the winter maximum sea ice extent (SIE) around Antarctica was interrupted by contraction, beginning in 2016 and continuing into 2019. This unexpected behavior motivates a closer look at factors controlling the position of the outer ice margin.We analyzed sea ice concentration (SIC) estimates derived from passive microwave sensors with differing resolutions (SSM/I, AMSR-E and AMSR2) to identify spatial and temporal statistics of the sea ice edge deVned by 15% SIC. The low-pass Vltered position of the ice edge is similar in different products, with the maximum northward position determined by proximity to the relatively warm waters of the Antarctic Circumpolar Current. Higher resolution SIC products reveal greater spatial detail along the convoluted margin, resulting in a relatively longer sea ice perimeter. Spectral analysis does not identify statistically signiVcant peaks in length scales along the margin; however, visual comparison with geostrophic velocities and sea surface temperature inferred from satellite altimetry suggests that advection of sea ice by mesoscale eddies is an important mechanism for deforming the ice edge in some regions, such as the Bellingshausen Sea. We analyze a high-resolution (dx=5 km), coupled ocean-sea ice model which realistically represents the annual expansion of sea ice to quantify the dynamic and thermodynamic roles of eddies in sea-ice mass balance and SIE. These eddy effects on the sea ice edge are not well represented in coarser-grid ocean reanalysis products such as ECCO-2, motivating an investigation of how to represent eddy/sea-ice interactions in global climate models.

Understanding the processes of wetland boundary morphodynamics is critical to evaluating the vulnerability of wetlands. Specifically, analysis of changes in sediment supply at the wetland boundary due to the effects of global sea-level rise and local subsidence provides a more comprehensive depiction of future changes in coastal wetlands. This requires analysis through numerical modeling along the wetlands area to provide predictive information for future scenarios. However, sediment transport and morphological dynamics in adjacent bays have not been incorporated into regional models of wetland evolution. The study investigates the short-term sediment transport processes in a highly resolved wetland boundary within Galveston Bay during cold front passages. In-situ measurements verified the hydrodynamic and waves conditions at the salt marsh boundary during the period of two cold front passages. The model showed that the circulation of sediment fluxes to Galveston Bay was increased during the cold front passage. In addition, extensive wetlands flooding caused by cold fronts significantly affected the sediment supply to the wetlands. The sea-level rise adaptation of the salt marsh platform was verified by comparing the baseline and relative sea level rise models.

Machine learning has the potential to automate the analysis of vast amounts of raw geophysical data, allowing scientists to monitor changes in key aspects of our climate such as cloud cover in real-time and at fine spatiotemporal scales. However, the lack of large labeled training datasets poses a significant barrier for effectively applying machine learning to these applications. Transfer learning, which involves first pretraining a neural network on an auxiliary “source” dataset and then finetuning on the “target” dataset, has been shown to improve accuracy for machine learning models trained on small datasets. Across prior work on machine learning for geophysical imaging, different choices are made about what data to pretrain on, and the impact of these choices on model performance is unclear. To address this, we systematically explore various settings of transfer learning for cloud classification, cloud segmentation, and aurora classification. We pretrain on different source datasets, including the large ImageNet dataset as well as smaller geophysical datasets that are more similar to the target datasets. We also experiment with multiple transfer learning steps where we pretrain on more than one source dataset. Despite the smaller source datasets’ similarity to the target datasets, we find that pretraining on the large, general-purpose ImageNet dataset yields significantly better results across all of our experiments. Transfer learning is especially effective for smaller target datasets, and in these cases, using multiple source datasets can give a marginal added benefit.

The lateral erosion of the salt marshes in Galveston Island wetlands was investigated through field UAV surveys. Using a GNSS receiver on the ground and a GNSS receiver attached to the UAV, highly accurate location information for the UAV images was derived through PPK analysis. Based on the collected UAV images with location information, Structure from Motion (SfM) by GNSS-supported aerial triangulation was conducted to generate orthomosaic images of wetlands. Drone surveys have been conducted seven times over a year, and the boundary erosion (represented by boundary retreat) for each period was calculated. The averaged error was less than 2 cm for the horizontal direction in the final orthomosaic images of wetlands. The average wave power, hypothesized to be a significant indicator of potential wetlands erosion, was calculated by investigating the wave climate of the survey area through a simplified numerical model. The relationship between wave power and erosion rate was then summarized. The average amount of erosion from 2019 DEC to 2020 MAR, where winds from north to south were dominant, was higher than from 2020 MAR to 2020 OCT. Much erosion was observed in a protruding part of the wetland. This is caused by waves coming from many different directions surrounding these parts. The erosion of this protrusion had a significant influence on the average erosion rate of the survey areas.

With the availability of satellite-based precipitation products, it is pertinent to develop methods to use these data products for the purpose of design of hydraulic structures. The satellite precipitation products play a vital role in ungauged locations or when information is required a catchment scale. Prior to such applications, the accuracy and uncertainty associated with the products have to be investigated. This study compares Intensity Duration Frequency (IDF) curves using the recent precipitation product Global Precipitation Measurement (GPM-IMERG V6) with ground-based gauge data over the southeastern part of India and quantifies the uncertainty associated. Further, for comparison, a bias-corrected dataset is used in the study to understand the implication of bias correction of the satellite product in the IDF generation. The spatial correlation between the satellite IDF and the gauge-based IDF improves significantly after bias correction and the value is as high as 0.75 for 2- 10 year return period. The bias between the satellite IDF and gauge IDF is low in the north part of the study region and is high in the southeastern part, prone to extreme rainfall. Further, a significant percentage of the satellite-based IDFs (with and without bias correction) lie inside the confidence interval of the gauge-based data.

In recent decades, the application of Digital Elevation Models (DEMs) has been widely used in various aspects such as land management and flood planning since it reflects the actual topographic characteristic on the Earth’s surface. However, obtaining a high-quality DEM is often quite challenging because it is time-consuming, costly, and often confidential. This study presents an innovative approach to derive an improved vertical accuracy of CARTOSAT 10m DEM by blending it with publicly available SRTM (Shuttle Radar Topography Mission) DEM using machine learning methods such as Genetic Programming (GP) and Artificial Neural Networks (ANN). SRTM-1 DEM and CARTOSAT DEM in India are applied to GP and ANN to generate improved vertical accuracy high-quality DEM. The results revealed that the proposed approach improves the vertical accuracy by considering the reference as Ground control Points (GCPs) elevation from Differential Global Positioning System (DGPS) survey data. A significant improvement of 47 and 35% generated DEMs in RMSE compared to the SRTM-1 and CARTOSAT, respectively.

Geomagnetic disturbance, or perturbations in ground-level magnetic field vectors relative to the quasi-static terrestrial main field, induces geoelectric fields at and below Earth’s surface. This leads to geomagnetically induced currents in high-voltage electric power systems that can interfere with their operations. The sparse geospatial distribution of reliable real time ground magnetometers is not presently adequate for accurate geoelectric field estimation using traditional interpolation techniques, or even more sophisticated inverse models (for example, a spherical elementary current system) alone. To address this shortcoming, we first generate multivariate statistics on a regular grid using state-of-the-art global magneto-hydrodynamics (MHD) simulations. These synthetic data strongly resemble real observations in a statistical sense, although they do not generally reproduce the detailed time evolution of observations due to poorly known MHD boundary conditions. However, these statistics can and are used to regress on sparse observations in order to fill in, or “impute”, the unobserved points on a regular grid over North America.

After protoplanets have acquired sufficient mass to open partial gaps in their natal protostellar disks, residual gas continues to diffuse onto some horseshoe streamlines under effect of viscous dissipation, and meander in and out of the planets’ Hill sphere. Inside the Hill sphere, the horseshoe streamlines intercept gas flow in circumplanetary disks. The host stars’ tidal perturbation induce a barrier across the converging streamlines’ interface. Viscous transfer of angular momentum across this tidal barrier determines the rate of mass diffusion from the horseshoe streamlines onto the circumplanetary disks, and eventually the accretion rate onto the protoplanets. We carry out a series of numerical simulations to test the influence of this tidal barrier on super-thermal planets. In weakly viscous disks, protoplanets’ accretion rate steeply decreases with their masses above the thermal limit. As their growth time scale exceeds the gas depletion time scale, their masses reach asymptotic values comparable to that of Jupiter. In relatively thick and strongly viscous disks, protoplanets’ asymptotic masses exceed several times that of Jupiter. Such massive protoplanets strongly excite the eccentricity of nearby horseshoe streamlines, destabilize orderly flow, substantially enhance the diffusion rate across the tidal barrier, and elevate their growth rate until their natal disk is severely depleted. Based on the upper fall-off in the observe mass distribution of known exoplanets, we suggest their natal disks had relatively low viscosity (α ∼ 10−3), modest thickness (H/R ∼ 0.03 − 0.05), and limited masses (comparable to that of minimum mass solar nebula model).

Cover crops may influence soil health and functioning. However, little is known about the role of belowground root architectural traits in linking cover crop diversity with rhizosphere soil ecosystem properties. We hypothesize that cover crop diversity may improve root traits, which in return, could influence its effects on essential indicators of soil physicochemical heterogeneity, such as the composition of soil aggregate-size classes and nutrients, the soil organic matter (SOM) and soil organic carbon (SOC) contents, and microbial communities. We studied the impact of cover plant diversity on root traits, soil properties and microbial communities. The four soil aggregate-size classes, such as large macro- (>2000μm), small macro- (<2000-500μm), meso- (<500-250 μm), and micro-aggregates (<250 μm) were separated by the dry sieving. Root traits such as surface area (cm2) and length (cm) were quantified by image analysis using Winrhizo. The soil nutrient, SOM, and SOC contents were determined by standard methods. Plant diversity improved productivity, root architectural traits, composition of soil aggregate-size classes and nutrients, SOM and SOC contents, composition and networking of microbial communities. Our results suggest that competition among plant roots in species-rich than poor communities may improve rhizosphere soil carbon storage, composition of soil aggregate-size classes and microbial communities.

Crop leaf angle is a crucial feature of plant architecture which influences photosynthetic efficiency and yield. Therefore high-throughput mapping of leaf angle is of extreme interest for both precision agriculture and crop breeding operations. In this study, we propose a UAV-based hybrid approach by combining a radiative transfer model (PROSAIL) and deep neural networks to estimate leaf angle from multi-angular hyperspectral and LiDAR data. PROSAIL can simulate canopy hyperspectral reflectance from a given list of parameters, where Average Leaf Inclination Angle (ALIA) is one of the canopy parameters. The goal is to develop a deep learning-based inversion function that takes UAV hyperspectral reflectance and other PROSAIL parameters as input and estimate ALIA of Maize as output. The other PROSAIL parameters will be estimated using several machine learning pipelines from UAV hyperspectral and LiDAR information. We also propose a multi-angular reflectance scheme where each image pixel will generate multiple simulated and observed reflectance from the overlapping regions using different angles (i.e., solar zenith angle, viewing zenith angle, and relative azimuth angle between the sun and sensor) involved in the PROSAIL simulation. An automated Python-based tool was developed that can calculate all three PROSAIL angles for a given hyperspectral data cube and generate the simulated reflectance for every vegetation pixels per experimental plot. Since the proposed method incorporates both crop information (i.e., PROSAIL) and a data-driven approach (i.e., deep learning), the method can be easily transferable for other study areas and crops, and it can rely on less ground truth data.

Predicting the composition of soybean seeds while the plants are growing in the field is very important to understand how different genotypes, field condition and environment influence different seed composition parameters. Knowing this information at global scale is even more important to understand the dynamics of food insecurity and the interaction of seed composition with global environmental changes. This study aims to develop a machine learning-based soybean seed composition model from the fusion of PlanetScope, Sentinel and Landsat satellite images. Although satellite images provide global coverage throughout the year, it suffers from coarser spatial resolution. However, PlanetScope provides four-band (i.e., red, green, blue, and near infrared) multispectral imageries at approximately 3m spatial resolution daily. Alternatively, Sentinel-2B and Landsat-8 have coarser spatial resolution (10 - 30m), they provide enriched spectral resolution. Therefore, the objectives of this study are to 1) fuse the PlanetScope image with corresponding Landsat and Sentinel images, 2) evaluate several machine learning algorithms (e.g., partial least squares, support vector machine, random forest, and deep neural network) to predict protein and oil content of soybean seeds from the fused satellite images. Two soybean fields were established in 2020 and 2021 at Bradford, MO to perform the experiment. Corresponding PlanetScope, Sentinel, and Landsat images were downloaded and processed for the entire growth seasons. Current results indicate that deep neural network provide the best performance in predicting both protein and oil content of soybean. Future step is to assess different fusion algorithms and predict seed composition at regional or global scale.