Atmospheric photochemistry is essential for simulating atmospheric composition, impacting air quality and climate change. However, conventional numerical schemes of photochemistry within atmospheric models are computationally expensive, leading to simplifications or omissions of critical processes in weather and climate models. Previous attempts to leverage artificial intelligence (AI) scheme to reduce computational costs have faced obstacles such as the curse of dimensionality and error propagation, and most have been limited to box models without coupling into numerical models. Here, we develop an innovative AI PhotoChemistry (AIPC) scheme coupled into an atmospheric model (WRF-Chem). With Multi-Head Self-Attention algorithm (MHSA), we simulate 74 chemical species and 229 reactions following the SAPRC-99 mechanism. This marks the first implementation of a sophisticated photochemical mechanism within one unified AI model, enabling fast, accurate, and stable simulations without needing individual AI model for each species as previous works. Comparative analysis reveals that the AIPC scheme outperforms previous AI schemes using Multi-Layer Perceptron and Residual Neural Network algorithms, offering superior accuracy and computational efficiency. Moreover, fine-tuning learning rate and broadening network width within the MHSA algorithm are more effective for improving the AIPC scheme’s performance than adjusting batch size or increasing network depth. When coupling AIPC into WRF-Chem, this hybrid model with both physics and AI schemes reproduces the spatiotemporal distributions of various species on monthly time scale, and achieves substantial speed enhancement with ~8 times faster than conventional scheme. This advancement lays the groundwork for future development of weather and climate models with sophisticated chemical processes.

The formation of Antarctic Bottom Water (AABW) is a key process in the global ocean circulation, but modelling the formation and downslope flow of AABW represents an ongoing challenge for ocean and climate models due to the high horizontal resolution required. Here, we assess the formation and export of AABW to the abyss and its sensitivity to horizontal model resolution in a circumpolar ocean-sea ice model available at horizontal resolutions of 1/10°, 1/20° and 1/40°. The AABW transport across the 1000 m isobath of the Antarctic continental slope increases by 27% with 1/20° resolution compared to 1/10°, but there is no further transport increase at 1/40° resolution. This resolution dependency is strongest in the Ross and Weddell Seas, the two most important regions of AABW formation. The higher AABW export at 1/20° compared to 1/10° resolution is due to formation of denser waters on the continental shelf and less diapycnal mixing during the downslope flow. This has effects downstream in the abyss of the Australian Antarctic Basin which is better ventilated in the 1/20° case. We conclude that a horizontal resolution of 1/20° is sufficient to simulate AABW formation and export, in agreement with theory of the downslope flow of dense plumes.

Kilometer grid-length simulations over a variety of different locations worldwide are used as training data for a deep-learning model designed to predict clouds in a global climate model. The inputs to the neural network are profiles of temperature, humidity and pressure from the high-resolution model, averaged to the scale of the climate model. The outputs are profiles of cloud fraction, liquid water content and ice water content. The high-resolution data is coarse-grained to a range of sizes, allowing the model to learn how the cloud formation depends on the size of the area being considered. The machine-learnt cloud cover and condensate scheme is coupled to a global climate model and used to run multi-year simulations where the clouds predicted by the neural-network are fully interacting with the rest of the model.

Hydrodynamic disconnectivity between surface water and groundwater is common in arid environments. It is also prone to affect shallow streams in wetter climate. Sediment layers with low permeability, owing to clogging, for instance, reduce hydrodynamic connectivity. The resistance of this clogging layer, results in unsaturated infiltration, which is characterized by the non-linear Richards equation. Either due to lack of field information, parsimony regarding computational resources or mere misunderstanding of the system, infiltration is often assumed saturated or drastically simplified in hydrogeological models. Here we show the existence of three simple generic asymptotic solutions to the unsaturated problem of vertical steady-state infiltration through a clogged profile, which we associate to three regimes: one dominated by the clogging layer, one by the underlying sediments and one depending on both layers. We also argue that infiltration rate roughly grows linearly with ponding depth. These observations motivate a refined definition of clogging potentially helpful in model selection as well as two novel approximate infiltration formulas. Our refined generic analytic framework is useful to better understand and formalize infiltration.

Woody plant encroachment is a global phenomenon, observed in many of the world's drylands. In those with shallow soils overlying karst geology, rock moisture can be an important source of water for the encroaching woody plants. This source can be particularly important for trees to maintain basic physiological functions during extended droughts, which are becoming more frequent and intense owing to climate change. However, our understanding of rock moisture dynamics in karst drylands undergoing woody plant encroachment is still limited because of the scarcity of direct measurements. In this study, we evaluated soil and rock moisture dynamics at a semiarid site in the Edwards Plateau region of Texas. Our measurements over the course of three years showed that in shallow upslope terrain, the dynamic water storage in bedrock was roughly twice that of soil, while in downslope terrain, the dynamic storage was largely restricted to the soil layer. Most of the bedrock storage gains occurred during the first year, after two major storm events of approximately 95 mm, and that storage was gradually depleted during the following two years, when precipitation was below average. Importantly, in upslope terrain we found substantially larger water storage under woody plants, which suggests that they not only can access and utilize rock moisture but also play a role in enhancing bedrock water storage capacity. These interconnected abilities can help woody plants survive extended droughts---a factor crucial for understanding their persistence and proliferation in the shallow soils of the Edwards Plateau and similar karst regions.

Seasonal hydrological dynamics have profound socio-economic implications for communities in the Ganges-Brahmaputra-Meghna (GBM) River basin. Climate change and El Niño-Southern Oscillation (ENSO) phase are known to impact extreme flood magnitude in GBM River, however how they affect seasonal flooding pattern is not revealed. Utilizing large ensemble climate data (comprising 6000 years of non-warming and warming climate scenarios) and the global hydrodynamic model CaMa-Flood, we assess the influence of climate change and ENSO on seasonal hydrological patterns specially focusing on maximum river flow. The quantitative effects of La Niña and El Niño are calculated utilizing the Fractional Attribution Risk (FAR) method, separately for non-warming and historical climate scenarios. We assess climate change’s impact on flooding by contrasting historical and non-warming climate conditions using the FAR method. Climate change has substantially increased the maximum river flow for all seasons. In the monsoon season, climate change amplifies the likelihood of flooding with a 10-year return period of 34%, 46%, and 31% at the Hardinge Bridge, Bahadurabad, and Bhairab Bazar gauge stations of the Ganges, Brahmaputra, and Meghna Rivers, respectively. The influence of ENSO still remains significant even with the influence of climate change. ENSO influence presents a nuanced picture, exhibiting variations both between seasons and across different rivers within the GBM basin. The relationship between ENSO and seasonal flood occurrence in the GBM basin can be effectively elucidated by the upward movement of moisture through vertical wind velocity, which serves as a large-scale controlling factor for flood variation.

The ocean’s biological carbon pump (BCP) plays a key role in global carbon cycling by transporting biologically-fixed carbon from the surface to the deep ocean. Prior analyses of the BCP in Earth System Model (ESM) simulations have typically evaluated particulate organic carbon (POC) flux at a fixed export depth horizon of 100 m. However, this overlooks spatial and temporal variations in the depth that sinking POC must penetrate to reach the mesopelagic, or to sequester carbon from the atmosphere on climate-relevant timescales. We use depth-resolved POC flux output from eight Coupled Model Intercomparison Project Phase 6 (CMIP6) ESMs to compare global and regional changes in POC flux at five export depth horizons ‒100 m, the base of the euphotic zone (EZ depth), the particle compensation depth (PCD), the maximum annual mixed layer depth (MLDmax), and 1000 m ‒ under the high-emissions scenario SSP5-8.5. We also examine the relationship among net primary production, export efficiency from the surface ocean, and transfer efficiency to depth in key regions of the ocean, identifying model- and region-specific variations in the mechanistic drivers of POC flux changes in the deep ocean. Globally and spatially, trends in POC flux magnitude and decline are similar at the four surface export depth horizons, and multimodel variability in POC flux change by 2100 is greatest at the 1000 m export depth horizon (+4% to -55%). This indicates the importance of improving model parameterizations of transfer efficiency and of POC flux to the deep ocean.

The turbulent static pressure spectrum Epp(kx) as a function of longitudinal wavenumber kx in the roughness sublayer of forested canopies is of interest to a plethora of problems such as pressure pumping from snow or forest floor. Long term static pressure measurements above a sub-arctic forested canopy during the winter and spring were collected and analyzed for three snow cover conditions: trees and ground covered with snow, trees are snow free but the ground is covered with snow, and snow free cover. In all three cases, it is shown that Epp(kx) obeys the attached eddy hypothesis at low wavenumbers (1/δ

The pressure-stimulated rock current (PSRC) is an extremely important petrophysical phenomenon and the catastrophic characterization of solid Earth, which attributed to several physical mechanisms including dislocation and positive holes (Ph). The micro-crack development coupling with the two mechanisms runs through the whole process of loading rock to failure, but how its development and type affect PSRC has not been well studied so far. The acoustic emissions (AE) and PSRC between the loaded part and free part were detected synchronously when bi-axially load dried diorite specimens to failure. This study revealed the remarkable characteristics of staged-variation of PSRC: a low-stable fluctuation in the elastic stage, accompanied with an instantaneous step-rising of average 26.6 nA; then exhibited stead fluctuation at the step until failure, accompanied with a negative pulse of several hundred nanoampere. The little PSRC in the elastic stage was mainly attributed to the Ph activated by tensile micro-cracks; while the step-rising large PSRC was attributed to the sudden rising of number and ratio of shear micro-crack at the moment of elastoplastic transition; after that, the PSRC was generated mainly by the coupling of continuous massive Ph activation and dislocation sliding resulting from shear micro-cracks. A quantitative expression was established to describe the PSRC dynamics. The results enriched and refined the mechanisms of PSRC, which provide key basis for engineering rock mass monitoring and critical status scrutinizing, and are of reference meanings for the monitoring of seismic electrical, magnetic and related radiation as well as the identification of earthquake precursors.

Some of the classical models of tropical cyclone intensification predict tropical cyclones to intensify up to a steady intensity, which depends on surface fluxes only, without any relevant role played by convective motions in the troposphere, typically assumed to have a moist adiabatic lapse rate. Simulations performed using the non-hydrostatic, high-resolution model SAM in idealized settings (rotating radiative-convective equilibrium on a doubly-periodic domain) show early intensification consistent with these theoretical expectations, but different intensity evolution, with the cyclone undergoing an oscillation in wind speed. This oscillation can be linked to feedbacks between the cyclone intensity and air buoyancy: convective heating, radiative heating, and mixing with warm low stratospheric air warm the mid and upper troposphere of the cyclone stabilizing the air column and thus reducing its intensity. After the intensity decay phase, mid and upper tropospheric cooling, mostly through cold advection from the surroundings, cooled by radiation, rebuilds CAPE and triggers a new intensification. These idealized simulations thus highlight the potentially important interactions between a tropical cyclone, its environment and radiation.

The nature of an apparently random rock deposit, such as regolith on an extraterrestrial body, is often evaluated in terms of the size distribution of the rock particles. Especially, the power-law index of the cumulative size-frequency distribution (CSFD), , is often used to describe the distribution. However, an value is typically obtained only from surface images, raising questions about its ability to reflect the size distribution of the entire deposit. In this study, we perform laboratory and numerical experiments to demonstrate that an value obtained from surface images alone does not necessarily represent that of the entire granular deposit. The value is only a reliable indicator of the size distribution of surface rocks, which can still be useful to gain insight into the evolutionary history of an asteroid’s surface. However, to fully address the internal information, direct observations of subsurface materials or sample analysis are required.

Large continental transform faults are thought to cross-cut the crust and extend into the lithospheric mantle. However, the strain distribution associated with these faults at mantle depths is not well understood. At the heart of this question is the rheology of the lithosphere: when stressed by displacement on a trans-crustal fault, does the uppermost mantle localize strain in a continuing narrow fault zone, or is that strain instead distributed in a shear zone that widens with increasing depth? This study uses harmonic decomposition of receiver functions to measure the spatial and depth distribution of seismic anisotropy, a proxy for viscous deformation, around the Alpine Fault in Aotearoa New Zealand. Anisotropy aligned with the fault is present in the lithosphere at least 100 km away from the fault trace, suggesting that the Alpine Fault shear zone widens at depth.

A document by Vrinda D Desai. Click on the document to view its contents.

The continental United States is well instrumented with facilities for mid-latitude upper atmosphere research that operate on a continuous basis. In addition, citizen scientists provide a wealth of information when unusual events occur. We combine ionospheric total electron content (TEC) data from distributed arrays of GNSS receivers, magnetometer chains, and auroral observations obtained by citizen scientists, to provide a detailed view of the intense auroral breakup and westward surge occurring at the peak of the 10-11 May 2024 extreme geomagnetic storm. Over a 20-min interval, vertical TEC (vTEC) increased at unusually low latitude (~45°) and rapidly expanded azimuthally across the continent. Individual receiver/satellite data sets indicate sharp bursts of greatly elevated of vTEC (~50 TECu). Intense red aurora was co-located with the leading edge of the equatorward and westward TEC enhancements, indicating that the large TEC enhancement was created by extremely intense low-energy precipitation during the rapid substorm breakup.

Climate feedbacks determine how much surface temperatures will eventually warm to balance anthropogenic radiative forcing, but remain difficult to constrain. The climate feedback due to some process X is defined as the partial derivative of outgoing radiation at the top of the atmosphere with respect to surface temperature following a change in X, λX=-∂Rout/TS|X, with total climate feedback a summation from all processes, λtotal=∑λX. Standard approaches evaluate climate feedbacks from finite temporal changes in surface temperatures and outgoing radiation, following observed or simulated perturbations to climate state. However, this introduces significant linear combination error (λtotal≠∑λX) when the applied perturbation is large enough to achieve a good signal-to-noise ratio. This study presents a new semi-empirical evaluation of non-cloud climate feedbacks, constrained instead by spatial variation in outgoing radiation and climate state. First, we observationally constrain functional relations for outgoing radiation over ocean and land in terms of surface temperature, pressure, relative humidity, the height of the tropopause, fractional clound amount and latitude. Then, these functional relations are differentiated with respect to surface temperature to calculate the climate feedbacks for infinitesimal perturbation, eliminating linear combination error at high signal-to-noise ratio. We find, when combined with a recent cloud feedback estimate, a present-day total climate feedback of -0.99 (-0.75 to -1.22 at 66% range) Wm-2K-1. Our method is independent of temporal variation approaches to evaluate climate feedback allowing Bayesian combination to further reduce uncertainty.

Breaking-wave induced bioluminescence is a critical component of the biogeochemical process in the ocean. Understanding bioluminescence is important for monitoring red tides caused by bioluminescent microorganisms. In this study, we present the first numerical effort to quantify bioluminescent light intensity based on high-fidelity direct numerical simulations of breaking waves and a quantitative bioluminescent model. The dynamics of breaking waves are extensively validated through comparison with existing studies. We find that the time-averaged and Lagrangian-averaged shear stress saturates as surface tension effects decrease and wave steepness increases. The spatial distribution of light intensity correlates with the wave crest overturning and air bubbles generated in plunging breakers. Furthermore, we observe that the maximum light intensity asymptotically approaches the emission of single cells, suggesting the potential for cost-effective prediction models in future studies.

• Caribbean reefs systems are highly efficient structures protecting coast from extreme waves • Infragravity waves generated within the surf zone become as high as sea-swell waves downstream the reef crest. • Wave-induced transport have profound consequences on the connectivity within the reef system.

Surface runoff and infiltrated water interact with dynamic landscape properties en route to the stream, ranging from vegetation and microbial activities to soil and geological attributes. Stream solute concentrations are highly variable and interconnected due to these interactions, flow paths, and residence times, and often exhibit hysteresis with flow. Significant unknowns remain about how point measurements of stream solute chemistry reflect interdependent hydrobiogeochemical and physical processes, and how signatures are encapsulated as nonlinear dynamical relationships between variables. We take a machine learning approach to understand and capture these dynamical relationships and improve predictions of solutes at short and long time scales. We introduce a physical process-based ”flow-gate” into an LSTM (long short-term memory) model, which enables the model to learn hysteresis behaviors if they exist. Further, we use information-theoretic metrics to detect how solutes are interdependent, and iteratively select source solutes that best predict a given target solute concentration. The ”flow-gate LSTM” model improves model predictions (RSME values decrease from 1% to 32%) relative to the standard LSTM model for all nine solutes included in the study. The predictive improvements from the flow-gate LSTM model highlight the importance of lagged concentration and discharge relationships for certain solutes. It also indicates a potential limitation in the traditional LSTM model approach since flow rates are always provided as input sources, but this information is not fully utilized. This work provides a starting point for a predictive understanding of geochemical interdependencies using machine-learning approaches and highlights potential improvements in model architecture.