We propose a multifaceted isoneutral eddy transport diagnostic framework that combines the stationary-transient and Leonard’s decomposition in large eddy simulation (LES). We diagnose the subfilter flux, the isotropic transport coefficient, and the anisotropic transport tensor or eigenvalues in the Southern Ocean (SO). The anisotropic tensor greatly reduces the reconstruction error of the subfilter flux because of its ability to distinguish the directionality of dynamic information, especially the topographic effect. A thorough analysis of the anisotropic tensor or transport eigenvalues reveals that the sign combination of the transport eigenvalues of the symmetric tensor links to the evolution of domainintegral large-scale PV enstrophy and the combination of different signs is most often, meaning the dominance of filamentation process in the SO. In the region with intense anisotropy, the dominant eigenvector tends to be perpendicular to the large-scale PV gradient, indicating an important role of the PV barrier mechanism in the SO transport process. The two distinct decompositions leveraged in our framework generate intriguing and profound results. Under the stationary-transient decomposition, we find a significant stationary contribution and the duality of the topographic effect which can not only anchors stationary structures but also organizes transient motions. Leonard’s decomposition, allows us to investigate the collective effects of the standing wave train, cross-scale interaction, and subfilter eddy-eddy interaction on the filtered space-time scale. We emphasize the complete subgrid flux, not the mere Reynolds term, and the LES framework needs to be considered in the subgrid parameterization of the coarse resolution ocean model.

....

The Underwater Vision Profiler (UVP) provides abundant in situ data of the marine particle size distribution (PSD) on global scales and has been used for a diversity of applications, but the uncertainty associated with its measurements has not been quantified. Here we use a global compilation of UVP (version 5) observations of the PSD to assess the sampling uncertainty associated with the UVP’s sampling characteristics. We model UVP sampling uncertainty using Bayesian Poisson statistics and provide formulae for the uncertainty associated with a given sampling volume and observed particle count. We also model PSD observations using a power law with an exponential cutoff to better match the low concentration associated with rare large particles as seen by the UVP. We use the two shape parameters from this statistical model to describe changes in the PSD shape across latitude band, season, and depth. The UVP sampling uncertainty propagates into an uncertainty for modeled carbon flux exceeding 50%. The statistical model is used to extend the size interval used in a PSD-derived carbon flux model, revealing a high sensitivity of the PSD-derived flux model to the inclusion of small particles (80-128 microns). We close with recommendations on how to revise the carbon flux model, and we provide avenues to address additional uncertainties associated with UVP-derived carbon flux calculations.

Recent observations from LOFAR indicate that multiple, spatially distributed corona bursts can occur in lightning processes on the order of 10 microseconds. The close proximity of the corona bursts in space and time poses a great observation challenge for dedicated lightning radio interferometers, typically with $<$100 m baselines. This paper reports simulations to show the interferometry results that would be obtained with a typical lightning interferometer for such a lightning process. In particular, spatially-separated corona bursts at fixed locations may be seen as a fast ($>10^7$ m/s) propagating source for an instrument with resolution greater than the spatial separation of the bursts. The implications and suggestions for lightning interferometry studies are discussed in the paper.

Identifying the heterogeneous conductivity field and reconstructing the contaminant release history are key aspects of subsurface remediation. Achieving these two goals with limited and noisy hydraulic head and concentration measurements is challenging. The obstacles include solving an inverse problem for high-dimensional parameters, and the high-computational cost needed for the repeated forward modeling. We use a convolutional adversarial autoencoder (CAAE) for the parameterization of the heterogeneous non-Gaussian conductivity field with a low-dimensional latent representation. Additionally, we trained a three-dimensional dense convolutional encoder-decoder (DenseED) network to serve as the forward surrogate for the flow and transport processes. Combining the CAAE and DenseED forward surrogate models, the ensemble smoother with multiple data assimilation (ESMDA) algorithm is used to sample from the Bayesian posterior distribution of the unknown parameters, forming a CAAE-DenseED-ESMDA inversion framework. We applied this CAAE-DenseED-ESMDA inversion framework in a three-dimensional contaminant source and conductivity field identification problem. A comparison of the inversion results from CAAE-ESMDA with physical flow and transport simulator and CAAE-DenseED-ESMDA is provided, showing that accurate reconstruction results were achieved with a much higher computational efficiency.

The Ross Ice Shelf (RIS) has long obscured the structure of the extended crust of the southern Ross Embayment. Here, we use airborne magnetics data from the ROSETTA-Ice project to estimate the depth to crystalline basement and sediment thickness beneath the RIS. We estimate the depth to the top of the magnetic crust, using a Werner deconvolution in a 2-D moving window applied to aeromagnetic data at 10 km line spacing. The result is then filtered, clustered, and gridded solutions to achieve a continuous basement surface. This method was tuned with seismic acoustic basement in the Ross Sea and magnetic basement determined from Operation Ice Bridge flight line data over both the Ross Sea and the RIS. Sub-RIS sediment thickness was then calculated relative to bathymetry. Shallow basement and thin sediments beneath the RIS define two major basement highs. A throughgoing, wide basement high is located midway across the RIS, and appears to form a southward continuation of the Ross Sea’s Central High. The other prominent basement feature underlies Roosevelt Island, with a continuation SE to the grounding zone. An elongate basin between the two highs deepens southward, reaching its greatest depth beneath Siple Dome. A deep oval basin flanks Crary Ice Rise. Both basins contain >2 km sediment. Bordering the Transantarctic Mountains (TAM) there is a deeper and broader basement basin that contains a narrow, linear, NW trending basement ridge. The magnetic basement is shallower toward Marie Byrd Land and deeper near the TAM/East Antarctica, consistent with contrasts in magnetic and gravity signatures of the crust on either side of the mid-RIS high. The trend of basement highs and basins parallel to the central TAM front suggests that the basement relief ≥2 km is structurally controlled, and is a product of regional extension. Basin sediments may be a source of deformable subglacial bed conditions for grounded ice, and basin flanking faults may control geothermal flux and subglacial water transport. These considerations have a bearing on sub-RIS boundary conditions that are important for cryosphere-ocean numerical modelling frameworks. Our work extends the current knowledge of sub-RIS sediment distribution, continental rifting orientations and the likely locations of basin flanking faults.

Soil and Water Assessment Tool (SWAT) is one of the widely used hydrological models, especially it has been successfully applied for the assessment of the impact of land use land cover and best management practices scenarios. But it is less applicable for type of research that requires integration or optimization with other models since it cannot update the land-use and best management practice information efficiently and it should be run separately when results for the multiple watersheds are needed. These days, the attention on the water security has been growing and interdisciplinary works desires integration of the hydrological model with other models are being highlighted. Thus, there are needs of development of the surrogate model which is computationally efficient and applicable to multiple watersheds. In this research, we propose the surrogate model of the SWAT with novel machine learning techniques such as random forest model. As a first step, the models are trained with the SWAT data from one watershed, which is a Maumee River Watersheds. Models for flow, mineral phosphorus, total nitrogen, total phosphorus, and sediment transport are built separately, and the model performance was above satisfactory level based on R-squared value, Nash-Sutcliffe efficiency, and percent bias. In addition, the surrogate models were tested for the different best management practices adoption scenarios and were trained additional data to make the model valid for the wide range of the best management practices adoption ratios. Finally, the surrogate models were expanded to multiple watersheds, by training SWAT results from Huron River Watersheds and River Raisin and they evaluated with the R-squared value. High R-squared values indicated that the surrogate model could be used in place of SWAT.

Soil erosion is the most common type of land degradation, and it has become a global environmental issue that reduces soil productivity and water quality. LULC change related with climatic and geomorphologic states of the area affects land degradation. This study aims at estimating the annual average soil erosion for the study area, Nagavali River Basin, and also focuses to analyze the impact of land use land cover change on the annual average soil erosion rates of the Nagavali river basin. In this study, the soil properties, elevation, and topography of the study are considered to be constant. This study was done by using the Revised Universal Soil Loss Equation (RUSLE) model. This equation includes factors like rainfall, soil, land cover, cultivation practices, and slope, for the estimation of soil erosion. Each layer of the factors affecting soil erosion was prepared and integrated using GIS techniques. By using the RUSLE model, the present study identified that the soil erosion rates for the years 1990 and 2020 ranged from 0 - 2364.46 t/ha/yr and 0 - 7857.21 t/ha/yr respectively for the Nagavali river basin. The LULC change analysis for the years 1990 and 2020 revealed that the erosion rate increased from 2364.46 t/ha/yr to 7857.21 t/ha/yr in the Nagavali river basin. The results also depicted that the area under very severe erosion class increased drastically from 1990 to 2020 while the area under very slight erosion class decreased from 1990 to 2020. This shows that LULC change has a significant impact on increase of soil erosion rates in Nagavali River Basin.

This study examines the impact of climate-sea level controls on the vegetation and evolution of the Niger Delta during the Late Quaternary. The extraneous controls on the environment outlined in this context confirm a direct link between vegetation dynamics (pollen data), sediment supply, and the landscape evolution of the Niger Delta between 20 ka and 6.5 ka. Two phases of sedimentation are recognized based on multiple proxies analyzed in three gravity cores obtained from the shallow offshore at ~40 m water depth. Phase I records abundant occurrences of Poaceae, Cyperaceae, and Podocarpus pollen from a dry hinterland, charred grass cuticles, nonmarine alga Pediastrum, high Ti/Zr ratio, and lower sedimentation from 20-11.7 ka. Phase II records an expansion of mangrove vegetation, high Fe/S ratio, and increase in planktonic foraminifera between 11.7 ka and 6.5 ka. This second phase is attributed to sea-level rise and higher sedimentation during the development of delta plain and mangrove vegetation on the gently sloping shelf. These sequential records provide a new clue about the link between the evolutionary stages of the Niger Delta landscape and vegetation dynamics during two distinct time-bound intervals, which potentially delineate the boundary between two Marine Isotope Stages: MIS2 (late glacial period) and MIS1 (interglacial period). Keyword: Sea level- climate change, Late Quaternary, mangrove pollen, biogeochemistry, Niger Delta, landscape evolution.

Hydropower plants and large storage reservoirs upstream of water-dependent power plants represent physical assets with a substantial role in the function of both river and power grid systems. Hydropower has the potential to play an important role in decarbonization strategies as a means to achieve reliability and resilience in an increasingly complex grid; for example, as a means to offset variability in other renewable resources and provide long term energy storage. The power grid and river systems are interlinked through hydropower plants, resulting in hydropower operators needing to balance potentially competing interests of water management, as well as power demands. Despite this multisectoral dependency, the representation of hydropower in power system resource adequacy studies typically differs from the representation in water management adaptation studies, especially at the regional scale. The objective of this presentation is to clarify the concept of hydropower flexibility (and adaptation) in a way that can bridge gaps between the tools used by and expectations of water managers, hydropower operators and power system operators. The developed qualitative multisystem, multiscale approach to hydropower flexibility can be leveraged to highlight the value, facilitate the compatibility and complementarity, and inform on the generalization of technology innovation and climate change adaptation strategies.

We perform a statistical study of 3-s ultra-low frequency (ULF) waves using MMS observations in the Earth’s foreshock region. The average phase velocity in the plasma rest frame is determined to be anti-sunward, and the intrinsic polarization is right-handed. We further examine the linear instability conditions based on drift-bi-Maxwellian distribution functions according to observed plasma conditions. The resulting instability is a solution to the common dispersion equation of the ion/ion right-hand non-resonant and left-hand resonant instabilities. The predicted wave propagation is also predominantly anti-sunward. The cyclotron resonant conditions of the solar wind and backstreaming beam ions are evaluated, and we find that in some cases, the anti-sunward propagating waves can be resonant with beam ions, which was overlooked in previous studies. The result suggests that the dispersion equation provides the 3-s ULF waves a fundamental explanation that unifies a rich variety of resonant conditions. In the later stage, the 3s ULF waves could further develop into Short Large Amplitude Magnetic Structures, contributing to the turbulence in the foreshock region.

Nitrous oxide (N2O) is one of the important greenhouse gases (GHGs), with its global warming potential 265 times greater than that of carbon dioxide (CO2). About 60% of the anthropogenic N2O emission is from agriculture production. To date, estimating N2O emissions from cropland remains a challenging task because the related microbial origin processes (e.g. incomplete nitrification and denitrification) are controlled by a diverse factors of climate, soil, plant and human activities. In this study, we developed a ML model with physical/biogeochemical domain knowledge, namely knowledge guided machine learning (KGML), for simulating daily N2O fluxes from the agriculture ecosystem. The Gated Recurrent Unit (GRU) was used as the basis to build the model structure. A range of ideas have been implemented to optimize the model performance, including 1) hierarchical structure based on variable causal relations, 2) intermediate variable (IMV) prediction and transfer, 3) inputting IMV initials for constraints, 4) model pretrain/retrain, and 5) multitask learning. The developed KGML was pre-trained by millions of synthetic data generated by an advanced PB model, ecosys, and then re-trained by observations from six mesocosm chambers during three growing seasons. Six other pure ML models were developed using the same data from mesocosm chambers to serve as the benchmark for the KGML model. The results show that KGML can always outperform the PB model in efficiency and ML models in prediction accuracy of capturing N2O flux magnitude and dynamics. Besides, the reasonable predictions of IMVs increase the interpretability of KGML. We believe the footprint of KGML development in this study will stimulate a new body of research on interpretable machine learning for biogeochemistry and other related geoscience processes.

Solar wind-magnetosphere coupling functions have been in use for almost 50 years. In that time, a very large number of formulations have been proposed. As they become increasingly subsumed into systems analysis and machine-learning studies of the magnetosphere, it is timely to establish best practice in their derivation and their limitations. This paper carried out a number of studies to establish some key points. Particular attention is paid to the best metric used to evaluate their performance and how it depends on the application for which the coupling function is intended.

Field survey, satellite image interpretation, geological map interpretation, literature review, GPlates reconstruction and LA-ICP-MS U-Pb dating of synkinematic calcites demonstrate that ~E/W-shortening in eastern Oman was significant and related to oblique convergence of Arabia and India from 32.5 to 20 Ma. Approximately N/S-oriented compressive structures, WNW to NNW-striking sinistral faults and ~E/W-oriented normal faults characterize a major shear zone in the eastern Oman Mountains (Hajar Shear Zone, HSZ) and its wrench-fault assemblage within an area spanning ~250 km x ~50 km. More than 10,000 mostly NW-striking lineaments as deduced from satellite image interpretation and numerous faults/folds indicate that strain of the HSZ is widely distributed but concentrated along WNW to NNW-striking major faults/fault zones at the SW margin of the Saih Hatat Dome. These faults/fault zones represent reactivated basement faults. GPlates reconstructions reveal that N-drifting India rotated 8° counter-clockwise with respect to fixed Arabia from 32.5 to 20 Ma, leading to ~100-135 km E/W-convergence between both plates (minimum value). This convergence created the sinistral HSZ with a displacement of a few to several tens of kilometers. Independently from the GPlates time constraints, two U-Pb ages of synkinematic calcites, crystallized along faults during HSZ movement, yield compatible ages of 30.08 ±0.47 and 22.31 ±2.15 Ma (2 standard error). E/W-shortening also affected the northern Oman Mountains, creating the ~N/S-oriented Hagab Thrust in the Musandam Peninsula and the Jabal Hafit Anticline.

Estimation of design flood is a crucial task in water resources engineering. Regional Flood Frequency Analysis is one of the widely used approaches for estimating design flood in ungauged basin. In the present research, we develop an eXtreme Gradient Boost based ML model for RFFA. The proposed approach relies on developing a regression model between flood quantiles and the commonly available catchment descriptors. In this study, the CAMELs data for 671 catchments from USA was used to study the efficiency of the approach. Further, the results were compared with the traditional methods such Multiple Linear Regression (MLR) and Artificial Neural Networks (ANN). XGB is a decision-tree-based ensemble machine learning algorithm that uses gradient boosting as a framework. The results revealed that the XGB based approach resulted in estimates with highest accuracy when using all the available catchment descriptors (i.e., mean annual rainfall(MAR), drainage area, fraction forest, mean annual potential evapotranspiration (MAPET), mean annual temperature, rainfall intensity, slope, fraction snow, soil porosity, and soil conductivity) both during training and validation. Four distinct models consisting of three to ten descriptors were examined for 2-, 5-, 10-, 25-, 50-, and 100-year return periods, all of the models exhibit smaller mean absolute error values and root mean square error values with percentage bias ranging from -10 to +10. A model with three predictor variables has comparable performance to other models. Drainage area, rainfall intensity, MAR, and fraction snow are the most efficient predictor variables, while MAPET, Slope, Temperature, Fraction Forest, Soil Porosity, and Soil Conductivity have low significance in predicting design flood for an ungauged catchment. The XGB modeling approach that has been proposed can be applied to different places throughout the world.

Flood frequency analysis assumes that annual peak flood events occur independently of each other, regardless of previous flood events (the independent and identically distributed (i.i.d.) assumption); however, annual peak flood records do not necessarily appear to conform to these assumptions. We tested the i.i.d assumption by analyzing the effects of the Pacific Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO) on 250 naturally flowing annual peak flood records across the entire western North American margin. Using permutation tests on quantile-quantile (Q-Q) plots, we found that the PDO has a greater impact on the magnitude of annual peak floods than the AMO. Twenty-six percent of the gauges have higher magnitude annual floods depending on the PDO phase (p < 0.1). Next, we examined the interacting effects of the PDO and AMO on the frequencies of lower and upper quartile annual peak floods, and found reinforcing, cancelling, and dominating effects. Lastly, we used permutation t-tests on the Julian dates of seasonal maximum and minimum streamflows to assess the impact of the PDO and AMO. We found that the PDO and AMO have substantial effects on the dates of winter maximum and summer minimum streamflow dates across the coastal margin. Since these two climate oscillations have significant effects on the magnitudes of annual peak floods, the i.i.d. assumption does not hold. Hence, we advocate for the need to re-assess baseline flood analysis in western North America to improve flood management strategies.

In this talk, we first examine how microphysical models of transient deformation might manifest at the macroscale which may be summarized as a distinct frequency dependence of dissipation that extends beyond typical models used within geophysics (e.g., Maxwell and Burger viscoelastic models). We will explore several seismological, geodetic, and geological observations that show a strong indication of the activation of transient deformation. These examples include seismic normal modes and tides that span minutes to decades, and observations of viscoelastic rebound in the response to melting ice sheets at time scales from weeks to thousands of years. For these, we will feature examples from Antarctica and Greenland. In all our cases we demonstrate that laboratory derived constitutive laws describing a full spectrum of deformation mechanisms can help to reconcile different inferences of viscoelastic structure.To arrive at these conclusions, theoretical and observational insights from a wide range of spatio-temporal data (the microphysical and planetary-scale) and disciplines (rock physics, seismology and geodynamics) have been combined. The talk will end with implications of the importance of considering transient deformation in other examples of Earth dynamics and associated challenges that remain.

Carbonyl sulfide (OCS) is a non-hygroscopic trace species in the free troposphere and the primary sulfur reservoir maintained by direct oceanic, geologic, biogenic and anthropogenic emissions and the oxidation of other sulfur-containing source species. It’s the largest source of sulfur transported to the stratosphere during volcanically quiescent periods. Data from 22 ground-based globally dispersed stations are used to derive trends in total and partial column OCS. Middle infrared spectral data are recorded by solar-viewing Fourier transform interferometers that are operated as part of the Network for the Detection of Atmospheric Composition Change between 1986 and 2020. Vertical information in the retrieved profiles provides analysis of discreet altitudinal regions. Trends are found to have well-defined inflection points. In two linear trend time periods ~2002 - 2008 and ~2008 - 2016, tropospheric trends range from ~0.0 to (1.55 ± 0.30 %/y) in contrast to the prior period where all tropospheric trends are negative. Regression analyses show strongest correlation in the free troposphere with anthropogenic emissions. Stratospheric trends in the period ~2008 - 2016 are positive up to (1.93 ± 0.26 %/y) except notably low latitude stations that have negative stratospheric trends. Since ~2016, all stations show a free tropospheric decrease to 2020. Stratospheric OCS is regressed with simultaneously measured N$_2$O to derive a trend accounting for dynamical variability. Stratospheric lifetimes are derived and range from (54.1 ± 9.7)y in the sub-tropics to (103.4 ± 18.3)y in Antarctica. These unique long-term measurements provide new and critical constraints on the global OCS budget.