Hyunje Yang
Compound Flood Lab.
Cockrell School of Engineering
The University of Texas at Austin
301 E Dean Keeton St, Austin, TX 78712
RESEARCH INTERESTS
Research vision: Toward transferable, probabilistic compound flood forecasting from site-specific emulation to regional-to-global early warning
Transferable AI Modeling
Geometry-aware neural operators that generalize flood prediction to unseen coastlines
Compound Coastal Flooding
Understanding how storm surge, rainfall, river discharge, and tide combine to flood the coast
Forecast Uncertainty and Early Warning
Uncertainty-aware flood forecasting for probabilistic early warning
Hydrology of Ungauged Systems
Machine-learning prediction of soil and catchment hydrology using terrain and remote sensing
RESEARCH EXPERIENCE
May 2026 – Jul. 2026
Oak Ridge National Laboratory (ORNL)
May 2026 – Jul. 2026
Graduate Research Intern · Oak Ridge, TN, USA
Jul. 2026 – Present
Research Collaborator · Austin, TX, USA (Remote)
Geometry-aware neural operators to scale storm-surge prediction from local coastal domains toward regional-to-global coverage on unstructured meshes
May 2026 – Present
- Developing geometry-aware neural operators that learn storm-surge dynamics and local geometry interaction on unstructured coastal meshes.
- Evaluating generalization across unseen coastal geometries to scale a unified deep learning framework toward regional-to-global coverage.
Investigating coastal wetland resistance and recovery using global tropical-cyclone records and satellite observations
May 2026 – Present
Developing a diffusion-based surrogate for reconstructing historical storm-surge dynamics
May 2026 – Present
Aug. 2023 – Present
The University of Texas at Austin (UT Austin)
Ph.D. Student, Civil, Architectural and Environmental Engineering · Austin, TX, USA
UT Austin Engineering Fellowship, Cockrell School of Engineering and UT Austin Graduate School
Mask-Aware Multimodal Deep Operator Network for Spatially Contiguous Mapping of Soil Water Retention Curves
Feb. 2026 – Present
- Developing a physically constrained DeepONet that integrates soil physicochemical and topographic information while accommodating missing input variables.
- Generating continuous soil water retention curves and spatially contiguous soil hydraulic property maps at 30 m resolution.
Texas-Scale Assessment of Multi-Driver Compoundedness in Coastal Flooding
Jan. 2026 – Present
- Quantifying the relative contributions and spatial patterns of storm surge, rainfall, and river discharge across historical storms.
- Identifying storm- and location-dependent shifts in dominant flood drivers.
Physics-Informed Machine Learning for Compound Flooding Prediction
May 2025 – Present
- Utilizing the SFINCS (Super-Fast INundation of CoastS) model to generate training data.
- Incorporating governing physics constraints, including mass and momentum conservation equations, into state-of-the-art deep learning models to improve physical consistency and predictive accuracy.
Integration of Machine Learning and Static Flood Models for Coastal Inundation Simulation
Aug. 2023 – Dec. 2025
- Developed a hybrid model (C1PK-Flood) integrating a machine learning-based storm surge prediction model (C1PKNet) and a static flood model (MatFlood) to rapidly generate high-resolution storm surge inundation maps.
- Proposed a novel method to optimize time series inputs of tropical cyclone parameters.
Reconstruction of Missing Data for Hurricane-Damaged Buildings using XGBoost
Aug. 2023 – Aug. 2024
- Developed an imputation model using structural, geospatial, hazard, and damage level data to reconstruct missing attributes of hurricane-damaged buildings.
- Assessed model performance across different regions and identified key features to support post-disaster building assessments.
Mar. 2018 – Apr. 2023
National Institute of Forest Science (NIFoS)
Researcher, Forest Environment and Conservation Department · Seoul, South Korea
[Government R&D Projects]
Establishment of Big Data and Integrated Utilization System for Flash Floods in Forest Watershed
Mar. 2021 – Apr. 2023
- Develop flash flood forecasting models using time series flood datasets and machine learning approaches.
- Find the optimal devices and algorithms for flash flood warning systems suited to forested areas.
Development of Forest Water Management Technology to Reduce Stream Depletion and Nonpoint Source Pollution
Mar. 2021 – Apr. 2023
- Monitor and collect data on changes in water quality after forest fire and forest thinning as the preliminary stage for analysis.
Quantification and Improvement of Forest Water Yield for Sustainable Supply of Freshwater
Mar. 2018 – Apr. 2023
- Calibrate interest parameters and analyze their uncertainties on physical hydrological simulation model via Bayesian approach and Generalized Likelihood Uncertainty Estimation (GLUE), using in situ measurement of the water cycle in different forest stands.
- Collect hydrological monitoring data for hydrograph analysis; compared various estimation methods to quantify the baseflow most rationally from streamflow in forested catchments.
- Estimate the soil hydraulic properties and predictive uncertainties based on a national scale spatial big dataset using machine learning models; examined the influence of environmental features on soil properties through sensitivity analysis.
※ Selected as a 10-year national R&D project with a $2.8M grant/year.
Long-term Monitoring of Flow Characteristics in Different Forest Stands and Locations
Mar. 2018 – Apr. 2023
- Collect the long-term water level and meteorological data from 52 sites and compare their hydrological characteristics.
- Evaluate the effect of forest thinning on the forest water cycle changes based on double mass curves and nonparametric statistics.
Jun. 2013 – Feb. 2018
Seoul National University (SNU)
Researcher, Department of Forest Environmental Science · Seoul, South Korea
Hydraulic Relation of Discharge and Velocity in Mountain Streams Using Salt Dilution Method
Mar. 2016 – Feb. 2018
- Measured in situ stream mean velocity and discharge via salt dilution; developed a mean velocity prediction formula based on forested topographical information and stream discharge using a nondimensionalized equation.
Monitoring of Floods in Mountainous Areas and Critical Runoffs in Urban Areas
Jun. 2013 – Feb. 2016
- Joined the 5th CALS Research Fellowship Program as an undergraduate researcher.
- Measured in situ streambed materials and organized flash flood velocity data in urban areas.
EDUCATION
| Aug. 2023 – May 2027 (Expected) |
The University of Texas at Austin (UT Austin) · Austin, TX, USA Ph.D. Candidate in Civil, Architectural and Environmental Engineering
|
| Mar. 2016 – Feb. 2018 |
Seoul National University (SNU) · Seoul, South Korea M.S. in Forest Environmental Science
|
| Mar. 2012 – Feb. 2016 |
Seoul National University (SNU) · Seoul, South Korea B.S. in Forest Environmental Science
|