Minho Kim

Minho Kim

PhD Candidate

UC Berkeley

I’m a PhD candidate in Environmental Planning at UC Berkeley under the guidance of Professor John Radke and Professor Marta Gonzalez. Currently, I am interested in using GeoAI (computer vision, deep learning, network data science) to model natural hazards and predict risks for environmental planning and management of complex urban and natural systems.

My ongoing research works include:

  • Data-driven fire spread models using machine learning (Funded by C3.AI)
  • Burn severity prediction and post-fire debris flow risk on critical infrastructure (Funded by Caltrans)

My previous research works include urban remote sensing, land cover mapping, image fusion, renewable energy forecasting, and deep learning for sustainable development of cities and the environment. I received my bachelor’s and master’s degree in Civil and Environmental Engineering at Seoul National University, where I was fortunate to be advised by Professor Yongil Kim and was part of the SPINS Lab reserach group. For my master’s thesis, I researched local climate zone classification using multispectral satellite images via multi-scale convolutional neural networks. For my bachelor’s thesis, I used DInSAR on Sentinel-1 images to detect North Korea’s nuclear test site.

Download my CV.

Interests
  • Environmental Modeling
  • GeoAI (Remote Sensing & GIScience)
  • Risk & Resilience of Complex Systems
  • Network Science
  • Machine learning
Education
  • PhD in Environmental Planning

    2021~, UC Berkeley

  • MS in Civil & Environmental Engineering

    2019~2021, Seoul National University

  • BSc in Civil & Environmental Engineering

    2012~2017, Seoul National University

Recent Work

1. Urban and Environmental Planning

  • [Research] Shared responsibility: A novel spatial metric for WUI planning and community resilience
  • [Paper] From local to micro: Exploratory data analysis on urban forests and microclimates in Portland, Oregon, USA (*Accepted at IGARSS 2023 as co-author)

2. Natural Hazards

  • [Research] Mapping post-fire debris flow risk using machine learning-based burn severity prediction
  • [Paper] Towards global open-source fire spread modeling using machine learning (In Preparation)

3. GeoAI and ML/DL

  • [Research] Wildfire burn severity prediction using deep learning
  • [Paper] Semantic segmentation of enhanced lifeform maps using high resolution satellite images (*Oral Presentation at IGARSS 2023)

Research

Publications and Conference Proceedings

Experience

Related research and work experience

 
 
 
 
 
Graduate Student Researcher
University of California, Berkeley
May 2023 – Present Berkeley, California
  • Project: Mapping post-fire soil burn severity and risk to critical transportation infrastructure (Funded by Caltrans)
 
 
 
 
 
Graduate Student Researcher
University of California, Berkeley
Jan 2022 – May 2022 Berkeley, California
  • Project: Multi-Scale Analysis for Improved Risk Assessment of Wildfires Facilitated by Data and Computation. (C3AI) Link
 
 
 
 
 
Graduate Researcher
Mar 2021 – Jul 2021 Seoul, South Korea
  • Research assistant at SPINS Lab working on multiple research projects :
    • Mapping Basic Spatial Information in North Korea Using Very High Resolution Satellite Images (The Institute for Peace and Unification Studies) Publication Link
    • Detection and Monitoring of Natural Disasters Using Multi-Modal and Multi-Sensor Remotely-Sensed Imagery (Ministry of Interior & Safety)
    • Development of an End-to-End Deep Learning based Technique to Generate Very High Resolution Environmental Data (National Research Foundation) Publication Link
 
 
 
 
 
Lab Manager
Jan 2021 – Jul 2021 Seoul, South Korea
  • Lab manager at SPINS Lab
 
 
 
 
 
Research Assistant
Mar 2019 – Feb 2021 Seoul, South Korea
  • Wildfire damage assessment using multispectral satellite and Planetscope images Publication Link
  • Solar power forecasting using geostationary satellite images and meteorological data Publication Link
  • Spatiotemporal fusion of multispectral and thermal satellite images Publication Link
  • Development of high-quality environmental data using satellite images via deep learning algorithms Publication Link
 
 
 
 
 
Ammunition Inspector
Republic of Korea Army
May 2017 – Jan 2019 Pocheon, South Korea
  • Recorded ammunition transactions and composed ammunition inventory reports using Excel
  • After working hours, self-studied satellite and radar remote sensing
  • Contributed to writeup on pan-sharpening image fusion research using Worldview images Publication Link
 
 
 
 
 
Research Assistant (Undergraduate Intern)
Aug 2016 – Feb 2017 Seoul, South Korea
  • Analyzed ground deformations in inaccessible, remote areas using Sentinel-1 SAR images
  • Assisted with fieldwork using a ground-based hyperspectral imager to monitor crop health
 
 
 
 
 
Research Assistant (Co-op Student)
Sep 2011 – Jan 2012 London, Canada
  • Researched on photoacoustic image reconstruction of a line source using multiple regularization percentages with the addition of maximum intensity projection using Matlab

Teaching

Instructor and teaching experience

 
 
 
 
 
Graduate Student Instructor
University of California, Berkeley
Jan 2024 – Present Berkeley, California
  • LDARCH/ESPM 289 Applied Remote Sensing
 
 
 
 
 
Lead Instructor
University of California, Berkeley
Aug 2022 – Dec 2021 Berkeley, California
  • GEOG/LDARCH C188 Geographic Information Systems
  • Teaching effectiveness: 6.311/7 (Dept. avg: 6.230/7)
 
 
 
 
 
Graduate Student Instructor
University of California, Berkeley
Aug 2021 – Dec 2021 Berkeley, California
  • GEOG/LDARCH C188 Geographic Information Systems
 
 
 
 
 
Teaching Assistant
Mar 2020 – Jul 2021 Seoul, South Korea
  • Graduate courses: Satellite Image Interpretation, Advanced Remote Sensing (KOMPSAT VHR imagery), Advanced Surveying
  • Undergraduate courses: Leadership for Civil Engineers, Spatial Informatics and Systems, Remote Sensing

Contact Me