arXiv subjects

Auroop R. Ganguly

Publications and source records attributed to Auroop R. Ganguly.

At least 19 recordsLinked to original sources

arXiv2026

Nonuniform and inequitable healthcare accessibility losses during flooding-amplified congestion across US cities

Cities in the United States and worldwide concentrate populations with diverse incomes, demographics and health risks, making access to healthcare facilities and emergency services critical. Urban flooding amplifies congestion and further degrades healthcare accessibility, but its impacts are uneven within and across cities, often burdening underserved populations and underplanned areas. These risks are intensifying as rapid development, aging infrastructure and climate change increase urban flood exposure. Across 30 of the most populous US cities, we show that peak-hour congestion alone reduces accessibility by 4% to 29%, while flooding further reduces it by 2% to more than 90%. These losses depend on flood type, transportation design, infrastructure planning, health vulnerability and spatial inequality. Flooding also increases spatial inequality in healthcare access, but the social burden varies by city and flood type, with urban coastal flooding disproportionately affecting less socially vulnerable populations in the United States. Our findings suggest that targeted approaches for embedding flood resilience in transportation and healthcare systems will be more effective than universal strategies, while allowing cities to learn from diverse urban regions facing similar accessibility risks.

physics.soc-ph↗

arXiv2026

Domain-Adaptive Climate Downscaling Under Temporal Distribution Shift

Deep-learning-based climate downscaling aims to learn relationships from historical low-resolution (LR) and high-resolution (HR) climate data to generate HR climate projections. However, this setting faces a temporal out-of-distribution (OOD) challenge: models trained on historical data are commonly applied to future projections whose distributions may differ substantially from the training period. This study investigates temporal OOD shift for daily temperature downscaling over the Continental United States using paired LR-HR model simulations. We propose a temporal domain-adaptive downscaling framework that combines supervised HR reconstruction on historical data with domain alignment between historical and future climate distributions. Experiments across future validation periods show that the proposed domain-adaptive model consistently outperforms statistical and deep-learning-based bias-correction methods, with the largest gains occurring when the temporal distribution shift is strongest. Spatial analyses indicate stronger improvements over high-elevation and topographically complex regions, along with higher spatiotemporal correlation with the HR target. The extreme analysis shows that domain adaptation also reduces upper-tail temperature bias relative to the non-adaptive model. These results demonstrate that temporal domain adaptation can improve the robustness of HR climate projections under non-stationary climate conditions.

cs.LG↗

arXiv2026

Intercomparison of Machine Learning Algorithms for Remote Sensing-based In-season Crop Mapping

In-season crop type mapping is critical for food security in the face of increasingly extreme climate-related threats to crops. Currently, the USDA Cropland Data Layer provides crop type labels at 30m resolution and is available the February after harvest, but no product exists that maps crop types before harvest with satisfactory accuracy that would allow emergency managers to respond to crop threats in near real time. Furthermore, the relative advantages of a wide range of algorithms have not been evaluated in a way that accounts for interannual variability, until this study. Here, Harmonized Landsat-Sentinel surface reflectance imagery time series and crop rotation history information are combined to map corn in Iowa and almonds in California at 30m resolution accurately by early June in unseen years, with robust quantification of uncertainty due to phenology and crop distribution. Thousands of model configurations across ten machine learning algorithms were compared using a year-wise cross-validation and a suite of metrics. Hyperparameter search revealed Support Vector Machines to be the most successful algorithm overall, with a mean F1 score of 0.74 (0.59) across five unseen validation years for almonds by early June in California (corn by early June in Iowa). Interannual variation was a large source of uncertainty, but patterns showed the potential to further improve performance with ensemble approaches or ancillary data. Future work may extend these methods to include multiclass maps of all crop types, CONUS-wide maps, and in-season crop yield forecasting.

cs.LG↗

arXiv2026

Learning Transferable Predictability Representations

We study the problem of assigning a scalar score to a short trajectory window that reflects its position on an ordered continuum of predictability regimes, spanning structured deterministic dynamics to unstructured stochastic noise. Existing methods address deterministic-versus-stochastic discrimination within a single system and do not produce scores with a consistent numerical interpretation across systems. We formalize this as ordinal estimation over a five-level predictability ladder and identify a structural source of cross-system ambiguity: ranking supervision alone leaves the score coordinate unfixed up to a monotone reparameterization, which we term the gauge freedom of ordinal scoring. We propose the Gauge-Fixed Ordinal Network (GON), a temporal convolutional model trained with an anchor-and-variance objective that pins level-wise score means to shared target coordinates. GON operates on 2-jet features that expose local trajectory geometry, preserved by smooth flows and disrupted by stochastic surrogate procedures. On five held-out dynamical systems, initializing from a pretrained GON checkpoint consistently outperforms training from scratch across all window budgets, with adaptation depth reflecting geometric proximity to the training family. Zero-shot scores retain ordinal structure at the stochastic boundary, where surrogate procedures most strongly disrupt nonlinear geometry, and pretrained initialization consistently beats scratch across all window budgets. Pairwise discrimination and globally coherent ordinal scoring are distinct properties requiring a stable score coordinate for cross-system transfer, with direct implications for predictability assessment, model selection, and early-warning diagnostics across natural and engineered dynamical systems.

cs.LG↗

arXiv2026

climt-paraformer: Stable Emulation of Convective Parameterization using a Temporal Memory-aware Transformer

Accurate representation of moist convective sub-grid-scale processes remains a major challenge in global climate models, as traditional parameterization schemes are both computationally expensive and difficult to scale. Neural network (NN) emulators offer a promising alternative by learning efficient mappings between atmospheric states and convective tendencies while retaining fidelity to the underlying physics. However, most existing NN-based parameterizations are memory-less and rely only on instantaneous inputs, even though convection evolves over time and depends on prior atmospheric states. Recent studies have begun to incorporate convective memory, but they often treat past states as independent features rather than modeling temporal dependencies explicitly. In this work, we develop a temporal memory-aware Transformer emulator for the Emanuel convective parameterization and evaluate it in a single-column climate model (SCM) under both offline and online configurations. The Transformer captures temporal correlations and nonlinear interactions across consecutive atmospheric states. Compared with baseline emulators, including a memory-less multilayer perceptron and a recurrent long short-term memory model, the Transformer achieves lower offline errors. Sensitivity analysis indicates that a memory length of approximately 100 minutes yields the best performance, whereas longer memory degrades performance. We further test the emulator in long-term coupled simulations and show that it remains stable over 10 years. Overall, this study demonstrates the importance of explicit temporal modeling for NN-based parameterizations.

physics.ao-ph↗

arXiv2026

What on Earth is AlphaEarth? Hierarchical structure and functional interpretability for global land cover

Geospatial foundation models generate high-dimensional embeddings that achieve strong predictive performance, yet their internal organization remains obscure, limiting their scientific use. Recent interpretability studies relate Google AlphaEarth Foundations (GAEF) embeddings to continuous environmental variables, but it is still unclear whether the embedding space exhibits a functional or hierarchical organization, in which some dimensions act as specialized representations while others encode shared or broader geospatial structure. In this work, we propose a functional interpretability framework that reverse-engineers the role of embedding dimensions by characterizing their contribution to land cover structure from observed classification behavior. The approach combines large-scale experimentation with a structural analysis of embedding-class relationships based on feature importance patterns and progressive ablation. Our results show that embedding dimensions exhibit consistent and non-uniform functional behavior, allowing them to be categorized along a hierarchical functional spectrum: specialist dimensions associated with specific land cover classes, low- and mid-generalist dimensions capturing shared characteristics between classes, and highgeneralist dimensions reflecting broader environmental gradients. Critically, we find that accurate land cover classification (98% of baseline performance) can be achieved using as few as 2 to 12 of the 64 available dimensions, depending on the class. This demonstrates substantial redundancy in the embedding space and offers a pathway toward significant reductions in computational cost. Together, these findings reveal that AlphaEarth embeddings are not only physically informative, but also functionally organized into a hierarchical structure, providing practical guidance for dimension selection in operational classification tasks.

cs.LG↗

arXiv2025

Land Cover Changes Cause Increased Losses during Photosynthetic Extremes

Human-induced carbon dioxide (CO2) emissions, primarily from fossil fuel combustion and changes in land use and land cover (LULCC), are a key contributor to climate change. As the climate warms, extreme events such as heatwaves, droughts, and wildfires have become more frequent and are projected to intensify throughout the 21st century. These escalating extremes are likely to further disrupt vegetation productivity, known as gross primary production (GPP), and reduce the ecosystem's capacity to absorb carbon. In this study, we employ a global Earth system model to assess how (a) CO2 emissions alone and (b) CO2 combined with LULCC influence the severity, frequency, and duration of GPP extremes. Our results show that negative GPP extremes periods of unexpectedly low carbon uptake are increasing more rapidly than positive extremes, especially under LULCC scenarios. The primary climate factor driving these extremes is soil moisture variability, which is influenced by fluctuations in both precipitation and temperature. The delayed responses of GPP to different climate drivers depend on the specific driver and geographical region. Overall, the highest incidence of GPP extremes arises from the combined influence of water stress, temperature anomalies, and fire-related disturbances.

physics.ao-ph↗

arXiv2025

Carbon Cycle Extremes Accelerate Weakening of the Land Carbon Sink in the Late 21st Century

Increasing surface temperature could lead to enhanced evaporation, reduced soil moisture availability, and more frequent droughts and heat waves. The spatiotemporal co-occurrence of such effects further drives extreme anomalies in vegetation productivity and net land carbon storage. However, the impacts of climate change on extremes in net biospheric production (NBP) over longer time periods are unknown. Using the percentile threshold on the probability distribution curve of NBP anomalies, we computed negative and positive extremes in NBP. Here we show that due to climate warming, about 88% of global regions will experience a larger magnitude of negative NBP extremes than positive NBP extremes toward the end of 2100, which accelerate the weakening of the land carbon sink. Our analysis indicates the frequency of negative extremes associated with declines in biospheric productivity was larger than positive extremes, especially in the tropics. While the overall impact of warming at high latitudes is expected to increase plant productivity and carbon uptake, high-temperature anomalies increasingly induce negative NBP extremes toward the end of the 21st century. Using regression analysis, we found soil moisture anomalies to be the most dominant individual driver of NBP extremes. The compound effect of hot, dry, and fire caused extremes at more than 50% of the total grid cells. The larger proportion of negative NBP extremes raises a concern about whether the Earth is capable of increasing vegetation production with growing human population and rising demand for plant material for food, fiber, fuel, and building materials. The increasing proportion of negative NBP extremes highlights the consequences of not only reduction in total carbon uptake capacity but also of conversion of land to a carbon source.

stat.AP↗

arXiv2025

Image Processing Techniques to Identify and Quantify Spatiotemporal Carbon Cycle Extremes

Rising atmospheric carbon dioxide due to human activities through fossil fuel emissions and land use changes have increased climate extremes such as heat waves and droughts that have led to and are expected to increase the occurrence of carbon cycle extremes. Carbon cycle extremes represent large anomalies in the carbon cycle that are associated with gains or losses in carbon uptake. Carbon cycle extremes could be continuous in space and time and cross political boundaries. Here, we present a methodology to identify large spatiotemporal extremes (STEs) in the terrestrial carbon cycle using image processing tools for feature detection. We characterized the STE events based on neighborhood structures that are three-dimensional adjacency matrices for the detection of spatiotemporal manifolds of carbon cycle extremes. We found that the area affected and carbon loss during negative carbon cycle extremes were consistent with continuous neighborhood structures. In the gross primary production data we used, 100 carbon cycle STEs accounted for more than 75\% of all the negative carbon cycle extremes. This paper presents a comparative analysis of the magnitude of carbon cycle STEs and attribution of those STEs to climate drivers as a function of neighborhood structures for two observational datasets and an Earth system model simulation.

stat.AP↗

arXiv2025

Predictions of damages from Atlantic tropical cyclones: a hierarchical Bayesian study on extremes

Bayesian hierarchical models are proposed for modeling tropical cyclone characteristics and their damage potential in the Atlantic basin. We model the joint probability distribution of tropical cyclone characteristics and their damage potential at two different temporal scales, while taking several climate indices into account. First, a predictive model for an entire season is developed that forecasts the number of cyclone events that will take place, the probability of each cyclone causing some amount of damage, and the monetized value of damages. Then, specific characteristics of individual cyclones are considered to predict the monetized value of the damage it will cause. Robustness studies are conducted and excellent prediction power is demonstrated across different data science models and evaluation techniques.

stat.AP↗

arXiv2025

Warming demands extensive tropical but minimal temperate management in plant-pollinator networks

Anthropogenic warming impacts ecological communities and disturbs species interactions, particularly in temperature sensitive plant pollinator networks. While previous assessments indicate that rising mean temperatures and shifting temporal variability universally elevate pollinator extinction risk, many studies often overlook how plant-pollinator networks of different ecoregions require distinct management approaches. Here, we integrate monthly near-surface temperature projections from various Shared Socioeconomic Pathways of CMIP6 Earth System Models with region-specific thermal performance parameters to simulate population dynamics in 11 plant pollinator networks across tropical, temperate, and Mediterranean ecosystems. Our results show that tropical networks, already near their thermal limits, face pronounced (50 percent) pollinator declines under high-emissions scenarios (SSP5-8.5). Multi-species management targeting keystone plants emerges as a critical strategy for stabilizing these high risk tropical systems, boosting both pollinator abundance and evenness. In contrast, temperate networks remain well below critical temperature thresholds, with minimal (5 percent) pollinator declines and negligible gains from any intensive management strategy. These findings challenge single-species models and uniform-parameter frameworks, which consistently underestimate tropical vulnerability while overestimating temperate risk. We demonstrate that explicitly incorporating complex network interactions, region-specific thermal tolerances, and targeted multi species interventions is vital for maintaining pollination services. By revealing when and where limited interventions suffice versus extensive management becomes indispensable, our study provides a clear blueprint for adaptive, ecosystem specific management under accelerating climate change.

q-bio.PE↗

arXiv2025

Network science disentangles internal climate variability in global spatial dependence structures

A comprehensive characterization of internal climate variability and irreducible uncertainty through initial-condition large ensembles of Earth system models across different spatiotemporal scales remains a significant challenge in climate science. In this study, we find significant differences in the spatial connectivity structures of temperature networks across ensemble members, with variations in long-range connections providing a distinguishing feature across the outcomes of initial conditions. Based on this, we introduce a novel quantifier, the 'Connectivity Ratio' (R), to encapsulate the spatial connectivity structure of each ensemble member by investigating the influence of internal climate variability on the global connectivity patterns in air temperatures. R allows us to characterize the variability of spatial dependence structure across the initial condition ensemble members as well as multiple models. Furthermore, we examine changes in spatial connectivity between near-term and long-term projections using R, which shows a potential shift in climate predictability under anthropogenic influence on a spatial scale.

physics.ao-ph↗

arXiv2024

Paraformer: Parameterization of Sub-grid Scale Processes Using Transformers

One of the major sources of uncertainty in the current generation of Global Climate Models (GCMs) is the representation of sub-grid scale physical processes. Over the years, a series of deep-learning-based parameterization schemes have been developed and tested on both idealized and real-geography GCMs. However, datasets on which previous deep-learning models were trained either contain limited variables or have low spatial-temporal coverage, which can not fully simulate the parameterization process. Additionally, these schemes rely on classical architectures while the latest attention mechanism used in Transformer models remains unexplored in this field. In this paper, we propose Paraformer, a "memory-aware" Transformer-based model on ClimSim, the largest dataset ever created for climate parameterization. Our results demonstrate that the proposed model successfully captures the complex non-linear dependencies in the sub-grid scale variables and outperforms classical deep-learning architectures. This work highlights the applicability of the attenuation mechanism in this field and provides valuable insights for developing future deep-learning-based climate parameterization schemes.

cs.LG↗

arXiv2024

Finer resolutions and targeted process representations in earth systems models improve hydrologic projections and hydroclimate impacts

Earth system models inform water policy and interventions, but knowledge gaps in hydrologic representations limit the credibility of projections and impacts assessments. The literature does not provide conclusive evidence that incorporating higher resolutions, comprehensive process models, and latest parameterization schemes, will result in improvements. We compare hydroclimate representations and runoff projections across two generations of Coupled Modeling Intercomparison Project (CMIP) models, specifically, CMIP5 and CMIP6, with gridded runoff from Global Runoff Reconstruction (GRUN) and ECMWF Reanalysis V5 (ERA5) as benchmarks. Our results show that systematic embedding of the best available process models and parameterizations, together with finer resolutions, improve runoff projections with uncertainty characterizations in 30 of the largest rivers worldwide in a mechanistically explainable manner. The more skillful CMIP6 models suggest that, following the mid-range SSP370 emissions scenario, 40% of the rivers will exhibit decreased runoff by 2100, impacting 260 million people.

physics.geo-ph↗

arXiv2024

Hybrid physics-AI outperforms numerical weather prediction for extreme precipitation nowcasting

Precipitation nowcasting, critical for flood emergency and river management, has remained challenging for decades, although recent developments in deep generative modeling (DGM) suggest the possibility of improvements. River management centers, such as the Tennessee Valley Authority, have been using Numerical Weather Prediction (NWP) models for nowcasting but have struggled with missed detections even from best-in-class NWP models. While decades of prior research achieved limited improvements beyond advection and localized evolution, recent attempts have shown progress from physics-free machine learning (ML) methods and even greater improvements from physics-embedded ML approaches. Developers of DGM for nowcasting have compared their approaches with optical flow (a variant of advection) and meteorologists' judgment but not with NWP models. Further, they have not conducted independent co-evaluations with water resources and river managers. Here, we show that the state-of-the-art physics-embedded deep generative model, specifically NowcastNet, outperforms the High-Resolution Rapid Refresh (HRRR) model, the latest generation of NWP, along with advection and persistence, especially for heavy precipitation events. For grid-cell extremes over 16 mm/h, NowcastNet demonstrated a median critical success index (CSI) of 0.30, compared with a median CSI of 0.04 for HRRR. However, despite hydrologically relevant improvements in point-by-point forecasts from NowcastNet, caveats include the overestimation of spatially aggregated precipitation over longer lead times. Our co-evaluation with ML developers, hydrologists, and river managers suggests the possibility of improved flood emergency response and hydropower management.

physics.ao-ph↗

arXiv2024

Topological Determinants of Resilience in Urban Rail Networks Facing Multi-Hazard Disruptions

This study examines the failure and recovery, two key components of resilience of nine major urban rail networks - Washington DC, Boston, Chicago, Delhi, Tokyo, Paris, Shanghai, London, and New York - against multi-hazard scenarios utilizing a quantitative approach focused on topological parameters to evaluate network resilience. Employing percolation-based network dismantling approach like Sequential Removal of Nodes and Giant Connected Component analysis, alongside random, centrality-based targeted attacks and flooding failure, findings reveal Domirank centrality's superior resilience in disruption and recovery phases. Kendall's tau coefficient's application further elucidates the relationships between network properties and resilience, underscoring larger networks' vulnerability yet faster recovery due to inherent redundancy and connectivity. Key attributes like average degree and path length consistently influence recovery effectiveness, while the clustering coefficient's positive correlation with recovery highlights the benefits of local interconnectivity. This analysis emphasizes the critical role of select nodes and the importance of balancing network design for enhanced resilience, offering insights for future urban rail system planning against multi-hazard threats.

physics.soc-ph↗

arXiv2023

A Tri-Level Optimization Model for Interdependent Infrastructure Network Resilience Against Compound Hazard Events

Resilient operation of interdependent infrastructures against compound hazard events is essential for maintaining societal well-being. To address consequence assessment challenges in this problem space, we propose a novel tri-level optimization model applied to a proof-of-concept case study with fuel distribution and transportation networks -- encompassing one realistic network; one fictitious, yet realistic network; as well as networks drawn from three synthetic distributions. Mathematically, our approach takes the form of a defender-attacker-defender (DAD) model -- a multi-agent tri-level optimization, comprised of a defender, attacker, and an operator acting in sequence. Here, our notional operator may choose proxy actions to operate an interdependent system comprised of fuel terminals and gas stations (functioning as supplies) and a transportation network with traffic flow (functioning as demand) to minimize unmet demand at gas stations. A notional attacker aims to hypothetically disrupt normal operations by reducing supply at the supply terminals, and the notional defender aims to identify best proxy defense policy options which include hardening supply terminals or allowing alternative distribution methods such as trucking reserve supplies. We solve our DAD formulation at a metropolitan scale and present practical defense policy insights against hypothetical compound hazards. We demonstrate the generalizability of our framework by presenting results for a realistic network; a fictitious, yet realistic network; as well as for three networks drawn from synthetic distributions. Additionally, we demonstrate the scalability of the framework by investigating runtime performance as a function of the network size. Steps for future research are also discussed.

eess.SY↗

arXiv2023

Explainable deep learning for insights in El Niño and river flows

The El Niño Southern Oscillation (ENSO) is a semi-periodic fluctuation in sea surface temperature (SST) over the tropical central and eastern Pacific Ocean that influences interannual variability in regional hydrology across the world through long-range dependence or teleconnections. Recent research has demonstrated the value of Deep Learning (DL) methods for improving ENSO prediction as well as Complex Networks (CN) for understanding teleconnections. However, gaps in predictive understanding of ENSO-driven river flows include the black box nature of DL, the use of simple ENSO indices to describe a complex phenomenon and translating DL-based ENSO predictions to river flow predictions. Here we show that eXplainable DL (XDL) methods, based on saliency maps, can extract interpretable predictive information contained in global SST and discover SST information regions and dependence structures relevant for river flows which, in tandem with climate network constructions, enable improved predictive understanding. Our results reveal additional information content in global SST beyond ENSO indices, develop understanding of how SSTs influence river flows, and generate improved river flow prediction, including uncertainty estimation. Observations, reanalysis data, and earth system model simulations are used to demonstrate the value of the XDL-CN based methods for future interannual and decadal scale climate projections.

physics.ao-ph↗