About

Guido Cervone, Ph.D., is a Professor of Geography, Meteorology and Atmospheric Science at The Pennsylvania State University. He serves as the Associate Director of the Institute for Computational and Data Sciences (ICDS) and is a faculty associate of the Earth and Environmental Systems Institute (EESI). In addition to his roles at Penn State, he is an Affiliate Scientist with the National Center for Atmospheric Research (NCAR) in Boulder, Colorado, and an Adjunct Professor at the Instituto Sant'Anna in Pisa, Italy. His research encompasses geoinformatics, machine learning, remote sensing, natural hazards, and renewable energy. Through these interdisciplinary fields, Professor Cervone contributes to advancing the understanding and application of computational and data science techniques to address environmental and atmospheric challenges.

Research topics

• Computer Science
• Data Mining
• Artificial Intelligence
• Geodesy
• Geology
• Seismology
• Geography
• Data science
• Virology
• Remote sensing
• Internal medicine
• Telecommunications
• Medicine

Selected publications

• MeltwaterBench: Deep learning for spatiotemporal downscaling of surface meltwater

  ArXiv.org · 2025-12-13

  preprintOpen access

  The Greenland ice sheet is melting at an accelerated rate due to processes that are not fully understood and hard to measure. The distribution of surface meltwater can help understand these processes and is observable through remote sensing, but current maps of meltwater face a trade-off: They are either high-resolution in time or space, but not both. We develop a deep learning model that creates gridded surface meltwater maps at daily 100m resolution by fusing data streams from remote sensing observations and physics-based models. In particular, we spatiotemporally downscale regional climate model (RCM) outputs using synthetic aperture radar (SAR), passive microwave (PMW), and a digital elevation model (DEM) over the Helheim Glacier in Eastern Greenland from 2017-2023. Using SAR-derived meltwater as "ground truth", we show that a deep learning-based method that fuses all data streams is over 10 percentage points more accurate over our study area than existing non deep learning-based approaches that only rely on a regional climate model (83% vs. 95% Acc.) or passive microwave observations (72% vs. 95% Acc.). Alternatively, creating a gridded product through a running window calculation with SAR data underestimates extreme melt events, but also achieves notable accuracy (90%) and does not rely on deep learning. We evaluate standard deep learning methods (UNet and DeepLabv3+), and publish our spatiotemporally aligned dataset as a benchmark, MeltwaterBench, for intercomparisons with more complex data-driven downscaling methods. The code and data are available at $\href{https://github.com/blutjens/hrmelt}{github.com/blutjens/hrmelt}$.

  PublisherOA PDFDOI

• Multimodal atmospheric super-resolution with deep generative models

  Machine Learning Earth · 2025-12-05 · 1 citations

  articleOpen access

  Diffusion models are a class of generative machine learning algorithms that can be used to sample from complex distributions. They achieve this by learning a score function, i.e. the gradient of the log-probability density of the data, and reversing a noising process using the same. Once trained, these diffusion models not only generate new samples but also enable zero-shot conditioning of the generated samples on observed data. This promises a novel paradigm for data and model fusion, wherein the implicitly learned distributions of pretrained diffusion models can be updated given the availability of online data in a Bayesian formulation. In this article, we apply such a concept to the super-resolution of a high-dimensional dynamical system, given the real-time availability of low-resolution (LR) and experimentally observed sparse measurements from multimodal data. Our experiments are performed for a super-resolution task that generates the ERA5 atmospheric dataset given sparse observations from a coarse-grained representation of the same and/or from unstructured experimental observations of the Integrated Global Radiosonde Archive radiosonde dataset. We also perform a data fusion task that leverages predictions from a data-driven atmospheric emulator as well. We discover that the generative model can balance the influence of multiple dataset modalities during spatiotemporal state reconstructions. Additional analysis on how score-based sampling can be used for uncertainty estimates is also provided.

  PublisherDOI

• GIScience in the era of Artificial Intelligence: a research agenda towards Autonomous GIS

  Annals of GIS · 2025-09-12 · 17 citations

  articleOpen access

  The advent of generative AI exemplified by large language models (LLMs) opens new ways to represent and compute geographic information and transcends the process of geographic knowledge production, driving geographic information systems (GIS) towards autonomous GIS. Leveraging LLMs as the decision core, autonomous GIS can independently generate and execute geoprocessing workflows to perform spatial analysis. In this vision paper, we further elaborate on the concept of autonomous GIS and present a conceptual framework that defines its five autonomous goals, five levels of autonomy, five core functions, and three operational scales. We demonstrate how autonomous GIS could perform geospatial data retrieval, spatial analysis, and map making with four proof-of-concept GIS agents. We conclude by identifying critical challenges and future research directions, including fine-tuning and self-growing decision-cores, autonomous modelling, and examining the societal and practical implications of autonomous GIS. By establishing the groundwork for a paradigm shift in GIScience, this paper envisions a future where GIS moves beyond traditional workflows to autonomously reason, derive, innovate, and advance geospatial solutions to pressing global challenges. Meanwhile, we emphasize that as we design and deploy increasingly intelligent geospatial systems, we carry a responsibility to ensure they are developed in socially responsible ways, serve the public good, and support the continued value of human geographic insight in an AI-augmented future.

  PublisherDOI

• Comparative Analysis between Multi-Scale Observation Product (MOP) with Flood2D-GPU Model and FEMA Products: A 2017 Hurricane Harvey Case Study

  2024-07-07

  article

  The study introduces a novel data fusion method that leverages diverse data sources for flood extent identification during emergencies, focusing on Hurricane Harvey. By analyzing various data sources, including satellite remote sensing, aerial photography, and ground observations from social media and municipal systems, spatial and temporal maps of inundation are generated for Harris County, TX, particularly Houston. These integrated data produce a multi-scale observation product (MOP) of cumulative inundation over time and space. The MOP is compared with three independent products: the Flood2D-GPU hydrologic model and two FEMA products for maximum inundation extent and building damage assessment.

  PublisherDOI

• Enhancing flood water level prediction through transfer learning with warmup learning rate scheduling in LSTM models: a comparative case study in Kentucky and Tennessee watersheds

  Modeling Earth Systems and Environment · 2024-12-19 · 2 citations

  articleSenior author
  PublisherDOI

• Integrating Thermal Infrared Imaging and Weather Data for Short-Term Prediction of Building Envelope Thermal Appearance

  Remote Sensing · 2024-10-26 · 5 citations

  articleOpen accessSenior author

  This study presents a novel deep-learning framework for predicting the thermal appearance of building envelopes under varying weather conditions based on a new dataset collected using a thermal infrared camera at 10 min intervals over a one-and-a-half-year period. Unlike existing studies that rely on simulated data or physical models that do not always accurately reflect the complex heat transfer processes in real buildings, we have collected a large dataset showing how a building behaves under different climatic conditions. We propose a novel deep-learning approach that integrates weather data and thermal imagery to predict the temperature distribution on the building façade for the next 24 and 48 h. The model uses a state-of-the-art recurrent neural network architecture, PredRNN V2, with an action conditioning mechanism to incorporate weather forecasting data into the prediction process. We evaluate this approach in terms of average accuracy, prediction accuracy in specific regions, and visual-perceptual performance of the images. The proposed framework achieves a prediction accuracy of 1.5 °C (root mean square error—RMSE) for the 24 h prediction and 2.04 °C (RMSE) for the 48 h prediction, outperforming baseline models in terms of temperature prediction accuracy and structural similarity of the predicted images.

  PublisherDOI

• Multiscale observation product (MOP) for temporal flood inundation mapping of the 2015 Dallas Texas Flood

  2024-06-21

  reportOpen access

  This paper presents a new data fusion multiscale observation product (MOP) for flood emergencies. The MOP was created by integrating multiple sources of contributed open-source data with traditional spaceborne remote sensing imagery to provide a sequence of high spatial and temporal resolution flood inundation maps. The study focuses on the 2015 Memorial Day floods that caused up to US$61 million of damage. The Hydraulic Engineering Center River Analysis System (HEC-RAS) model was used to simulate water surfaces for the northern part of the Trinity River in Dallas, using reservoir surcharge releases and topographic data provided by the US Army Corps of Engineers. A measure of fit assessment is performed on the MOP flood maps with the HEC-RAS simulated flood inundation output to quantify spatial differences. Estimating possible flood inundation using individual datasets that vary spatially and temporally allow an understanding of how much each observational dataset contributes to the overall water estimation. Results show that water surfaces estimated by MOP are comparable with the simulated output for the duration of the flood event. Additionally, contributed data, such as Civil Air Patrol, although they may be geographically sparse, become an important data source when fused with other observation data.

  PublisherOA PDFDOI

• Using human mobility data to detect evacuation patterns in hurricane Ian

  Annals of GIS · 2024-04-14 · 16 citations

  articleOpen accessSenior author

  Hurricane Ian in 2022 was a destructive category 4 Atlantic hurricane striking the state of Florida, which caused hundreds of deaths and injuries, catastrophic property damage, and an economic loss of more than $112 billion. Before the landfall of Ian in Florida, the state government issued evacuation orders in high-risk zones to reduce casualties and injuries. However, there is limited data available to monitor the actual evacuation patterns and compliance with the evacuation orders at a large geographic scale. This study utilizes human mobility data (i.e. SafeGraph Weekly Pattern) to analyse the spatial patterns of evacuation during Hurricane Ian in 2022. The objectives of the study include three key aspects: 1) proposing an analytical workflow that utilizes human mobility data to detect mobility patterns in disasters and other emergency events; 2) identifying significant evacuation patterns, and 3) revealing the spatial variations in the compliance with evacuation orders in the affected areas. Using data science and spatial analysis techniques, this study detected notable changes in population movements, both within Florida and nationwide, which are potentially linked to the hurricane-induced population evacuation. The distance decay pattern of population flows from Florida demonstrates a propensity for individuals to relocate to nearby areas during the hurricane. Furthermore, the increase in population outflows from the impacted areas suggests the effectiveness of mandatory evacuation orders. A more pronounced increase in outflows from designated mandatory evacuation areas points to the public awareness of the evacuation zone designation. This study provides large-scale, fine-resolution analysis of evacuation behaviours in natural disasters which cannot be easily detected in traditional data sources. The analytical workflows provide actionable tools for government agencies and policymakers to evaluate the effectiveness of evacuation orders and improve evacuation plans in future disasters.

  PublisherOA PDFDOI

• Arctic accessibility: recent trend in observed ship tracks and validation of arctic transport accessibility model

  Annals of GIS · 2024-07-21 · 4 citations

  articleOpen access

  The Arctic region is undergoing significant changes in maritime accessibility. This study investigates observed ship trajectories from 2013 to 2020 to demonstrate the recent trends of Arctic traffic. A notable surge in maritime activities has been observed, particularly during summer months, driven by economic interests and the increasing popularity of existing routes. Unique patterns in the northern Barents Sea have been observed where ships favour different routes based on seasonal ice conditions. Another contribution from this work is the validation of the Arctic Traffic Accessibility Model (ATAM) using the observed ship traffic data. Results show that the ATAM model underestimates the accessibility and vessel travel speed. This is largely due to outdated model parameters. The predefined ice multipliers and calculation of ice numerals used in the ATAM may not accurately reflect real-world conditions.

  PublisherDOI

• Geospatial big data: theory, methods, and applications

  Annals of GIS · 2024-10-01 · 13 citations

  articleOpen accessSenior author
  PublisherOA PDFDOI

Recent grants

• EarthCube Building Blocks: Collaborative Proposal: The Power of Many: Ensemble Toolkit for Earth Sciences

  NSF · $391k · 2016–2020

Frequent coauthors

• M. Kafatos

  Chapman University

  37 shared

• Weiming Hu

  Scripps Institution of Oceanography

  36 shared

• Luca Delle Monache

  University of California, San Diego

  25 shared

• Carolynne Hultquist

  University of Canterbury

  24 shared

• Nigel Waters

  23 shared

• R. P. Singh

  21 shared

• Pasquale Franzese

  University of Naples Federico II

  15 shared

• Elena Sava

  12 shared

Education

• Ph.D., Computational Sciences and Informatics

  George Mason University

  2005

• M.S., Computer Science

  George Mason University

  2000

• B.S., Computer Science

  The Catholic University of America School of Engineering

  1998