Searcharxiv⌕ Search

arXiv subjects

Sai Ravela

Publications and source records attributed to Sai Ravela.

At least 19 recordsLinked to original sources

Physics-Spawned Neural Integration of Rotating Shallow-Water Dynamics

Equation-derived neural integration can execute prescribed dynamics without training, while informative ensemble learning can estimate uncertain parameters and identify governing interactions without requiring differentiability. Together, these approaches motivate Physics-Spawned Informative Networks ($Φ$SpIN), a framework for prescribed and adaptive hybrid dynamical systems. As a first step toward flood modeling applications, we present $Φ$SpIN-SW, its prescribed-physics realization for the rotating shallow-water equations. Compatible C-grid operators on spherical Voronoi meshes define a computational graph that retains non-polynomial operations and advances the state through recurrent neural integration compiled with JAX. Analytical benchmarks characterize numerical errors, and global and coastal experiments closely reproduce reference solutions generated using the Model Prediction Across Scales (MPAS) model. At approximately 14~km global resolution, A100, H200 and B200 GPUs attain 9-, 19- and 34-fold higher timestep throughput than a 96-core MPAS-SW CPU reference under matched workloads. These results extend training-free neural integration beyond polynomial dynamics to complex geophysical simulation, establishing a computational foundation for subsequent adjoint-free ensemble estimation of physical and neural parameters and informative adaptation of operator structure.

physics.comp-ph↗

The Physics Is the Network: Fast, Training-Free Neural Emulation

Neural emulation need not require learning a substitute for known physics. We investigate PolyNets that couple a continuous-time neural representation of polynomial dynamics with a simple recurrent network for discrete-time integration. When the governing coefficients are known, the resulting tensor graph requires no training and preserves the prescribed numerical scheme while enabling efficient hardware execution and parallel simulation. Using the one-dimensional Korteweg-de Vries equation, we compare PolyNet with CPU and GPU finite-difference Runge-Kutta solvers and three trained neural models: a state-map emulator, a physics-informed predictor, and a learned-tendency integrator. Analytical soliton solutions provide a reference for amplitude and phase errors across grid resolutions, 10^6 dispersively scaled time steps, ensembles of up to 10^4 trajectories, and varying initial amplitudes. PolyNet achieves accuracy comparable to the numerical solvers, including at amplitudes outside the learned models' training conditions. It combines this fidelity with high GPU throughput, approximately constant over the tested resolution range, and favorable ensemble scaling.

cs.LG↗

The Neuro-Physical Inverter: A Modular Framework for Magnetotelluric Inversion Coupling Ensemble Conditioning with Residual Learning

We present the Neuro-Physical Inverter (NPI), a modular, uncertainty-aware framework for geophysical inversion that couples ensemble-based conditioning with constrained residual learning, demonstrated in the 1D magnetotelluric (MT) setting as a controlled testbed. The framework operates in two stages. An Ensemble-Conditional Gaussian Process (EnsCGP) conditions a prior ensemble of resistivity models on the observed response, producing a physically admissible reference ensemble. A residual-learning neural network then predicts targeted corrections to this reference, trained on synthetic data and fine-tuned per station for field application through a physics-coupled objective. Because an ensemble is conditioned, refined, and propagated through both stages, every estimate carries an associated ensemble spread. Synthetic experiments show that NPI systematically reduces ensemble-mean error without destabilizing the ensemble. Applied to broadband MT data from the Gabbs Valley geothermal region (Nevada, USA), NPI reduces the across-station mean misfit over the mid-period band while retaining comparable ensemble spread. The propagated ensemble yields a factor of uncertainty that serves as an operational measure of constraint within the assumed model class. Both stages are dimension-agnostic in formulation, and the design principles established here are intended to scale to higher-dimensional parameterizations.

cs.LG↗

Learning a Stochastic Differential Equation Model of Tropical Cyclone Intensification from Reanalysis and Observational Data

Tropical cyclones are among the most consequential weather hazards, yet estimates of their risk are limited by the relatively short historical record. To extend these records, researchers often generate large ensembles of synthetic storms using simplified models of cyclone intensification. Developing such models, however, has traditionally required substantial theoretical effort. Here we explore whether equation-discovery methods, a class of data-driven techniques designed to infer governing equations, can accelerate the process of developing simplified intensification models. Using observational storm data (IBTrACS) together with environmental conditions from reanalysis (ERA5), we learn a compact stochastic differential equation describing tropical cyclone intensity evolution. We focus on TCs because their dynamics are well studied and a hierarchy of reduced-order models exist, enabling direct comparison of the learned model to physically-derived counterparts. We find that the learned model simulates synthetic TCs whose intensification statistics and hazard estimates are consistent with observations and competitive with a leading physics-based TC intensification model. Our model also reproduces known nonlinear dynamical behavior of tropical cyclones, including as a saddle node bifurcation as inner core ventilation is increased. This result shows that equation-discovery approaches, when applied directly to storm intensity, can recover not only realistic statistics but also physically meaningful dynamical structure. These findings highlight the potential for data-driven methods to complement existing theory and reduced-order models in the study of extreme weather.

cs.LG↗

A Dynamic Learning Observatory Reveals the Rapid Salinization of Satkhira, Bangladesh

Soil salinity is a major environmental challenge in coastal Bangladesh, threatening agricultural productivity and local livelihoods. This study develops a machine-learning-based framework to predict and map soil salinity in Satkhira district by integrating field observations with Landsat-derived spectral indices. A total of 205 soil samples collected during 2024-2025 were used to train an Extreme Gradient Boosting (XGBoost) model, and predictions were further improved using a Generalized Additive Model (GAM). Spatial cross-validation was applied to reduce autocorrelation bias, and bootstrap resampling was used to quantify prediction uncertainty. The results show strong spatial variability of soil salinity, with higher concentrations in the southern and central coastal regions and lower levels in the northern inland areas. Vegetation indices, particularly NDVI, along with salinity-related spectral indicators, were identified as key predictors. 10-year-window peak-exposure maps generated for 2014-2023 reveal recurrent high-salinity zones and a persistent, expanding footprint of moderate-to-high salinity exposure across the central parts of the district. Uncertainty analysis indicates higher variability in coastal zones and improved prediction stability when multi-year datasets are combined. The proposed framework provides a robust and scalable approach for long-term monitoring of soil salinity. It supports climate-resilient agriculture, land-use planning, and evidence-based decision-making in coastal Bangladesh.

physics.geo-ph↗

Ensemble-Conditional Gaussian Processes (Ens-CGP): Representation, Geometry, and Inference

We formulate Ensemble-Conditional Gaussian Processes (Ens-CGP), a finite-dimensional synthesis that centers ensemble-based inference on the conditional Gaussian law. Conditional Gaussian processes (CGP) arise directly from Gaussian processes under conditioning and, in linear-Gaussian settings, define the full posterior distribution for a Gaussian prior and linear observations. Classical Kalman filtering is a recursive algorithm that computes this same conditional law under dynamical assumptions; the conditional Gaussian law itself is therefore the underlying representational object, while the filter is one computational realization. In this sense, CGP provides the probabilistic foundation for Kalman-type methods as well as equivalent formulations as a strictly convex quadratic program (MAP estimation), RKHS-regularized regression, and classical regularization. Ens-CGP is the ensemble instantiation of this object, obtained by treating empirical ensemble moments as a (possibly low-rank) Gaussian prior and performing exact conditioning. By separating representation (GP -> CGP -> Ens-CGP) from computation (Kalman filters, EnKF variants, and iterative ensemble schemes), the framework links an earlier-established representational foundation for inference to ensemble-derived priors and clarifies the relationships among probabilistic, variational, and ensemble perspectives.

math.ST↗

LASSE: Learning Active Sampling for Storm Tide Extremes in Non-Stationary Climate Regimes

Identifying tropical cyclones that generate destructive storm tides for risk assessment, such as from large downscaled storm catalogs for climate studies, is often intractable because it entails many expensive Monte Carlo hydrodynamic simulations. Here, we show that surrogate models are promising from accuracy, recall, and precision perspectives, and they "generalize" to novel climate scenarios. We then present an informative online learning approach to rapidly search for extreme storm tide-producing cyclones using only a few hydrodynamic simulations. Starting from a minimal subset of TCs with detailed storm tide hydrodynamic simulations, a surrogate model selects informative data to retrain online and iteratively improves its predictions of damaging TCs. Results on an extensive catalog of downscaled TCs indicate 100% precision in retrieving rare destructive storms using less than 20% of the simulations as training. The informative sampling approach is efficient, scalable to large storm catalogs, and generalizable to climate scenarios.

physics.ao-ph↗

Bangladesh's Amplified Coastal Storm Tide Hazard from Tropical Cyclones and Rising Sea Levels in a Warming Climate

The risk of extreme storm tides to Bangladesh's low-lying and densely populated coastal regions, already vulnerable to tropical cyclones, remains poorly quantified under a warming climate. Here, using a statistical-physical downscaling approach, our multimodel large-ensemble projections under the IPCC6 SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios show that Bangladesh's 100-year storm tide will likely intensify from 3.5 m to between 4.9 m and 5.4 m by the end of the 21st century. The Meghna-North Chattogram region is the most vulnerable, and the storm tide season will broaden significantly, amplifying the strongest during the late monsoon and late post-monsoon seasons. We project substantial increases in seasonal storm tide frequencies, with a four-fold increase in back-to-back extremes in the post-monsoon season. Across the SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios assessed using multiple climate models, the frequency of storm tide from destructive cyclones like Bhola and Gorky will significantly increase by 7-18 times and 6-23 times, respectively. Our study indicates a need to re-examine the ongoing coastal improvement and heighten the urgency to enhance coastal resilience in Bangladesh.

physics.ao-ph↗

A Root-Zone Soil Salinity Observatory for Coastal Southwest Bangladesh

The research assesses soil salinity in the southwest coastal region of Bangladesh, collecting a total of 162 topsoil samples between March 1 and March 9, 2024, and processing them following the standard operating procedure for soil electrical conductivity (soil/water, 1:5). Electrical conductivity (EC) measurements obtained using a HI-6321 advanced conductivity benchtop meter were analyzed and visualized using bubble density mapping and the Empirical Bayesian Kriging interpolation method. The findings indicate that soil salinity in the study area ranges from 0.05 to 9.09 mS/cm, with the highest levels observed near Debhata and Koyra. A gradient of increasing soil salinity is clearly evident from the northern to southern regions. This dataset provides a critical resource for soil salinity-related research in the region, offering valuable insights to support decision-makers in understanding and mitigating the impacts of soil salinity in Bangladesh's coastal areas.

physics.geo-ph↗

Rapid Climate Model Downscaling to Assess Risk of Extreme Rainfall in Bangladesh in a Warming Climate

As climate change drives an increase in global extremes, it is critical for Bangladesh, a nation highly vulnerable to these impacts, to assess future risks for effective adaptation and mitigation planning. Downscaling coarse-resolution climate models to finer scales is essential for accurately evaluating the risk of extremes. In this study, we apply our downscaling method, which integrates data, physics, and machine learning, to quantify the risks of extreme precipitation in Bangladesh. The proposed approach successfully captures the observed spatial patterns and risks of extreme rainfall in the current climate while generating risk and uncertainty estimates by efficiently downscaling multiple models under future climate scenarios. Our findings project that the risk of extreme rainfall rises across the country, with the most significant increases in the northeastern hilly and southeastern coastal areas. Projections show that the daily maximum rainfall for a 100-year return period could increase by approximately 50 mm/day by mid-century and around 100 mm/day by the end of the century. However, substantial uncertainties remain due to variations across multiple climate models and scenarios.

physics.ao-ph↗

Rapid Statistical-Physical Adversarial Downscaling Reveals Bangladesh's Rising Rainfall Risk in a Warming Climate

In Bangladesh, a nation vulnerable to climate change, accurately quantifying the risk of extreme weather events is crucial for planning effective adaptation and mitigation strategies. Downscaling coarse climate model projections to finer resolutions is key in improving risk and uncertainty assessments. This work develops a new approach to rainfall downscaling by integrating statistics, physics, and machine learning and applies it to assess Bangladesh's extreme rainfall risk. Our method successfully captures the observed spatial pattern and risks associated with extreme rainfall in the present climate. It also produces uncertainty estimates by rapidly downscaling multiple models in a future climate scenario(s). Our analysis reveals that the risk of extreme rainfall is projected to increase throughout Bangladesh mid-century, with the highest risk in the northeast. The daily maximum rainfall at a 100-year return period is expected to rise by approximately 50 mm per day. However, using multiple climate models also indicates considerable uncertainty in the projected risk.

physics.ao-ph↗

A Stochastic Compound Failure Model for Testing Resilience of Autonomous Fixed-Wing Aircraft I: Formulation and Simulation

This paper presents a Markov chain model to dynamically emulate the effects of adverse (failure) flight conditions on fixed-wing, autonomous aircraft system actuators. It implements a PX4 Autopilot flight stack module that perturbs the attitude control inputs to the plane's actuator mixer. We apply this approach in simulation on a fixed-wing autonomous aircraft to test the controller response to stochastic compound failures on a range of turning radii. Statistical measures of the differences between target and simulated flight paths demonstrate that a well-tuned PID controller remains competitive with adaptive control in a cascading, compound, transient failure regime.

eess.SY↗

Fully fluorinated non-carbon compounds NF3 and SF6 as ideal technosignature gases

Waste gas products from technological civilizations may accumulate in an exoplanet atmosphere to detectable levels. We propose nitrogen trifluoride (NF3) and sulfur hexafluoride (SF6) as ideal technosignature gases. Earth life avoids producing or using any N-F or S-F bond-containing molecules and makes no fully fluorinated molecules with any element. NF3 and SF6 may be universal technosignatures owing to their special industrial properties, which unlike biosignature gases, are not species-dependent. Other key relevant qualities of NF3 and SF6 are: their extremely low water solubility, unique spectral features, and long atmospheric lifetimes. NF3 has no non-human sources and was absent from Earth's pre-industrial atmosphere. SF6 is released in only tiny amounts from fluorine-containing minerals, and is likely produced in only trivial amounts by volcanic eruptions. We propose a strategy to rule out SF6's abiotic source by simultaneous observations of SiF4, which is released by volcanoes in an order of magnitude higher abundance than SF6. Other fully fluorinated human-made molecules are of interest, but their chemical and spectral properties are unavailable. We summarize why life on Earth-and perhaps life elsewhere-avoids using F. We caution, however, that we cannot definitively disentangle an alien biochemistry byproduct from a technosignature gas.

astro-ph.EP↗

Multirotor Ensemble Model Predictive Control I: Simulation Experiments

Nonlinear receding horizon model predictive control is a powerful approach to controlling nonlinear dynamical systems. However, typical approaches that use the Jacobian, adjoint, and forward-backward passes may lose fidelity and efficacy for highly nonlinear problems. Here, we develop an Ensemble Model Predictive Control (EMPC) approach wherein the forward model remains fully nonlinear, and an ensemble-represented Gaussian process performs the backward calculations to determine optimal gains for the initial time. EMPC admits black box, possible non-differentiable models, simulations are executable in parallel over long horizons, and control is uncertainty quantifying and applicable to stochastic settings. We construct the EMPC for terminal control and regulation problems and apply it to the control of a quadrotor in a simulated, identical-twin study. Results suggest that the easily implemented approach is promising and amenable to controlling autonomous robotic systems with added state/parameter estimation and parallel computing.

eess.SY↗

Learning Inter-Annual Flood Loss Risk Models From Historical Flood Insurance Claims and Extreme Rainfall Data

Flooding is one of the most disastrous natural hazards, responsible for substantial economic losses. A predictive model for flood-induced financial damages is useful for many applications such as climate change adaptation planning and insurance underwriting. This research assesses the predictive capability of regressors constructed on the National Flood Insurance Program (NFIP) dataset using neural networks (Conditional Generative Adversarial Networks), decision trees (Extreme Gradient Boosting), and kernel-based regressors (Gaussian Process). The assessment highlights the most informative predictors for regression. The distribution for claims amount inference is modeled with a Burr distribution permitting the introduction of a bias correction scheme and increasing the regressor's predictive capability. Aiming to study the interaction with physical variables, we incorporate Daymet rainfall estimation to NFIP as an additional predictor. A study on the coastal counties in the eight US South-West states resulted in an $R^2=0.807$. Further analysis of 11 counties with a significant number of claims in the NFIP dataset reveals that Extreme Gradient Boosting provides the best results, that bias correction significantly improves the similarity with the reference distribution, and that the rainfall predictor strengthens the regressor performance.

cs.LG↗

Downscaling Extreme Rainfall Using Physical-Statistical Generative Adversarial Learning

Modeling the risk of extreme weather events in a changing climate is essential for developing effective adaptation and mitigation strategies. Although the available low-resolution climate models capture different scenarios, accurate risk assessment for mitigation and adaption often demands detail that they typically cannot resolve. Here, we develop a dynamic data-driven downscaling (super-resolution) method that incorporates physics and statistics in a generative framework to learn the fine-scale spatial details of rainfall. Our method transforms coarse-resolution ($0.25^{\circ} \times 0.25^{\circ}$) climate model outputs into high-resolution ($0.01^{\circ} \times 0.01^{\circ}$) rainfall fields while efficaciously quantifying uncertainty. Results indicate that the downscaled rainfall fields closely match observed spatial fields and their risk distributions.

physics.ao-ph↗

An Adaptive Digital Autopilot for Fixed-Wing Aircraft with Actuator Faults

This paper develops an adaptive digital autopilot for a fixed-wing aircraft and compares its performance with a fixed-gain autopilot. The adaptive digital autopilot is constructed by augmenting the autopilot architecture implemented in PX4 flight stack with adaptive digital control laws that are updated using the retrospective cost adaptive control algorithm. In order to investigate the performance of the adaptive digital autopilot, the default gains of the fixed-gain autopilot are scaled down to degrade its performance. This scenario provides a venue for determining the ability of the adaptive digital autopilot to compensate for the detuned fixed-gain autopilot. Next, the performance of the adaptive autopilot is examined under failure conditions by simulating a scenario where one of the control surfaces is assumed to be stuck at an unknown angular position. The adaptive digital autopilot is tested in simulation, and the resulting performance improvements are examined.

eess.SY↗

Neural Networks as Geometric Chaotic Maps

The use of artificial neural networks as models of chaotic dynamics has been rapidly expanding. Still, a theoretical understanding of how neural networks learn chaos is lacking. Here, we employ a geometric perspective to show that neural networks can efficiently model chaotic dynamics by becoming structurally chaotic themselves. We first confirm neural network's efficiency in emulating chaos by showing that a parsimonious neural network trained only on few data points can reconstruct strange attractors, extrapolate outside training data boundaries, and accurately predict local divergence rates. We then posit that the trained network's map comprises sequential geometric stretching, rotation, and compression operations. These geometric operations indicate topological mixing and chaos, explaining why neural networks are naturally suitable to emulate chaotic dynamics.

cs.LG↗