Giha Lee

Giha Lee

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Giha Lee has a total of 91 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Seasonal variation monitoring of fluvial scene using PlanetScope data and machine learning method

    by Linh Nguyen Van, Lam Nguyen Van, May Thi-Tuyet Do, Mingeun Song, Giha Lee

    1. Rivers are four-dimensional systems whose channels, bars, and riparian vegetation shift in response to changes in flow magnitude, sediment supply, and seasonal hydrology. Flood pulses scour and ...


  • Open Access

    Tree-Based Regressor Comparison for Burn Severity Mapping: Spatially Blocked Validation Within and Across Fires

    by Linh Nguyen Van, Giha Lee

    Accurate, timely maps of post-fire burn severity are vital for rehabilitation, hydrologic hazard assessment, and ecosystem recovery in the western United States, where large, frequent wildfires and steep environmental gradients challenge model generalization. Machine learning models, particularly tree-based regressors, are increasingly used to relate satellite-derived spectral features to ground-based severity metrics such as the Composite Burn Index (CBI). However, model generalization across spatial domains, both within and between wildfires, remains poorly characterized. In this study, we benchmarked six tree-based regression models (Decision Tree-DT, Random Forest-RF, Extra Trees-ET, Bagging, Gradient Boosting-GB, and AdaBoost-AB) for predicting wildfire severity from Landsat surface reflectance data across ten U.S. fire events. Two spatial validation strategies were applied: (i) within-fire spatial generalization via Leave-One-Cluster-Out (LOCO) and (ii) cross-fire transfer via Leave-One-Fire-Out (LOFO). Performance is assessed with R2, RMSE, and MAE under identical predictors and default hyperparameters. Results indicate that, under LOCO, variance-reduction ensembles lead: RF attains R2 = 0.679, MAE = 0.397, RMSE = 0.516, with ET statistically comparable (R2 = 0.673, MAE = 0.393, RMSE = 0.518), and Bagging close behind (R2 = 0.668, MAE = 0.402, RMSE = 0.525). Under LOFO, ET transfers best (R2 = 0.616, MAE = 0.450, RMSE = 0.571), followed by GB (R2 = 0.564, MAE = 0.479, RMSE = 0.606) and RF (R2 = 0.543, MAE = 0.490, RMSE = 0.621). These results indicate that tree ensembles, especially ET and RF, are competitive under minimal tuning for rapid severity mapping; in practice, RF is a strong choice for an individual fire with local calibration, whereas ET is preferred when model transferability to unseen fires is paramount.


  • Open Access

    Investigating the Relationship Between Topographic Variables and Wildfire Burn Severity

    by Linh Nguyen Van, Giha Lee

    Wildfire behavior and post-fire effects are strongly modulated by terrain, yet the relative influence of individual topographic factors on burn severity remains incompletely quantified at landscape scales. The Composite Burn Index (CBI) provides a field-calibrated measure of severity, but large-area analyses have been hampered by limited plot density and cumbersome data extraction workflows. In this study, we paired 6150 CBI plots from 234 U.S. wildfire events (1994–2017) with 30 m SRTM DEM, extracting mean elevation, slope, and compass aspect within a 90 m buffer around each plot to minimize geolocation noise. Topographic variables were grouped into ecologically meaningful classes—six elevation belts (≤500 m to >2500 m), six slope bins (≤5° to >25°), and eight aspect octants—and their relationships with CBI were evaluated using Tukey HSD post hoc comparisons. Our findings show that all three factors exerted highly significant influences on severity (p < 0.001): mean CBI peaked in the 1500–2000 m belt (0.42 higher than lowlands), rose almost monotonically with steepness to slopes > 20° (0.37 higher than <5°), and was greatest on east- and northwest-facing slopes (0.19 higher than south-facing aspects). Further analysis revealed that burn severity emerges from strongly context-dependent synergies among elevation, slope, and aspect, rather than from simple additive effects. By demonstrating a rapid, reproducible workflow for terrain-aware severity assessment entirely within GEE, the study provides both methodological guidance and actionable insights for fuel-management planning, risk mapping, and post-fire restoration prioritization.


  • Open Access

    Rapid Urban Flood Detection Using PlanetScope Imagery and Thresholding Methods

    by Linh Nguyen Van, Giang V. Nguyen, Younghun Kim, May T. T. Do, Seongcheon Kwon, Jinhyeong Lee, Giha Lee

    With advances in optical satellite remote sensing, urban flood mapping (UFM) leveraging water’s distinct spectral characteristics for water identification is preferred and has gained more attention. PlanetScope’s daily 3 m resolution imagery enables detailed and time-sensitive flood monitoring. Unlike machine learning, which requires extensive training data, thresholding methods offer a faster and more adaptable solution for binary classification. Three global (Yen’s, Otsu’s, Isodata) and three local (Niblack, Sauvola, Gonzalez) thresholding methods, with their parameters optimized for each case study, were assessed in this study. Additionally, a hybrid approach was proposed and evaluated. In this approach, local thresholds are computed for each pixel, using the respective local thresholding method. Then, a global threshold is derived by calculating the simple arithmetic mean of all these local thresholds. This global threshold is subsequently applied across the entire image to perform a binary classification, distinguishing flooded from non-flooded areas. To enhance water detection, we also evaluated 26 remote sensing indices. Each was computed using two formulations—the normalized difference and the ratio—where at least one of the eight PlanetScope bands was NIR or RedEdge to enhance water detection. We tested this methodology on three flooding events with different water coverage scenarios in Brazil (34% water coverage), the USA (11%), and Australia (21%). The model performance was validated using the Matthews correlation coefficient (MCC) and the Fowlkes–Mallows index (FMI). The results demonstrated that combining PlanetScope imagery with carefully selected remote sensing indices and thresholding techniques enhances efficient UFM. The hybrid methods outperformed the others by capturing local variations while maintaining global consistency, with the MCC and the FMI exceeding 0.9. The indices incorporating NIR and RedEdge, particularly NDRE, achieved the highest accuracy. However, each flood event was best classified by a different combination of method and index, indicating that it is important to carefully select the appropriate remote sensing indices and thresholding techniques for each specific case. This framework provides a fast, effective solution for UFM, adaptable to diverse urban environments and flood conditions.


  • Open Access

    Uncertainty Evaluation of Deep Learning Models Using an Artificial Rainfall

    by Younghun Kim, Giha Lee

    Accurate rainfall prediction is essential not only for water resource management but also for forecasting and mitigating the impacts of climate change-driven weather events such as floods and droughts. Due to the high spatiotemporal variability of complex meteorological phenomena like rainfall, effective prediction necessitates in high-quality data collection, model application, and uncertainty analysis. Unlike existing studies that focus primarily on developing deep learning models to improve rainfall prediction accuracy, this study evaluates the uncertainty of rainfall predictions using pre-existing deep learning models, U-Net and ConvLSTM, with artificially generated elliptical rainfall data. Artificial rainfall data were designed with four temporal patterns: constant, gradually increasing, gradually decreasing, and time-varying. These patterns were applied in horizontal, vertical, and diagonal movements to evaluate the models' ability to handle spatiotemporal complexity. The results indicate that both deep learning models exhibited spatial smoothing issues on rainfall predictions over time. However, the U-Net model demonstrated superior spatiotemporal performance compared to ConvLSTM. While this study focuses solely on deep learning models for rainfall prediction, future research will consider factors such as data complexity and loss functions to conduct a comprehensive evaluation of prediction uncertainty. This work is expected to contribute to the development of methodologies for rainfall modeling using deep learning approaches. Funding: This research was supported by Disaster-Safety Platform Technology Development Program of the National Research Foundation of Korea(NRF) funded by the Ministry of Science and ICT. (No. 2022M3D7A1090338)

  • Open Access

    Simulation of debris flow in wildfire areas using the Nays2DFlood model

    by Chanul Choi, Giha Lee

    Wildfire-affected areas are increasingly prone to debris flows due to vegetation loss and changes in soil characteristics caused by intense heat. These conditions can lead to severe erosion, slope instability, and the mobilization of sediment during rainfall events, posing significant threats to downstream communities and infrastructure. Understanding debris flow dynamics in wildfire-prone regions is crucial for disaster risk reduction and effective management strategies.This study employs the Nays2DFlood model, a two-dimensional simulation tool, to investigate debris flow behavior in wildfire-damaged areas. To replicate realistic flow scenarios, the model integrates key datasets, including high-resolution Digital Elevation Models (DEMs), soil physical properties, and rainfall data. Through this simulation, we aim to identify critical factors influencing debris flow initiation and propagation in post-wildfire environments. Furthermore, the study provides an initial framework to analyze differences in debris flow behavior before and after wildfires, offering insights into the long-term impacts of vegetation loss on sediment transport and hydrodynamics.Preliminary results demonstrate that post-wildfire debris flows exhibit higher flow velocities and sediment transport rates than pre-wildfire conditions. These findings underline the importance of incorporating wildfire-induced changes into hazard prediction models. Additionally, the research highlights the potential risk to critical infrastructure such as dams, reservoirs, and downstream settlements, emphasizing the need for early warning systems and mitigation measures.This research is a foundational step for future comparative studies on pre- and post-wildfire debris flow dynamics. By advancing our understanding of these processes, this study contributes to the development of more accurate predictive tools and disaster management strategies, ultimately enhancing resilience in wildfire-prone regions. The insights gained will support policymakers and engineers in designing robust infrastructure and implementing effective land management practices to mitigate risks associated with wildfire-induced debris flows.Funding : This work is financially supported by Korea Ministry of Environment(MOE) as 「Research and Development on the Technology for Securing the Water Resources Stability in Response to Future Change(RS-2024-00332494)