Global Flood Susceptibility Map (GFSM v1)
An open, globally harmonized 30-meter flood susceptibility dataset built from multi-source Earth observation data and approximately 30.45 million training samples.
Mirza Waleed
I build GeoAI systems for
Flood intelligence, climate adaptation, geospatial machine learning, and earth observation operations.
I study how changing landscapes shape the risks people face. As a computational geographer, I use satellite observations and machine learning to understand floods, environmental change, and resilience.
01.
Selected Projects
Selected GeoAI systems and open-source tools for flood risk, land-cover intelligence, and urban climate analysis — including GFSM v1, a global flood susceptibility dataset, and BAM, a near-real-time wildfire mapping framework.
BAM: Self-Supervised Burn Area Mapping
BAM is a revolutionary, self-supervised machine learning framework designed for near-real-time mapping of wildfire burned areas globally. It bridges physics-based spectral analysis with data-driven ML refinement to eliminate reliance on manual training data, achieving robust 30-meter resolution mapping that resolves sub-pixel heterogeneity across diverse ecosystems and topographies.
Flood Risk Analytics for Pakistan
High-resolution (30m) flood susceptibility mapping and population exposure analysis for Pakistan, integrating ensemble machine learning with geospatial big data to support national disaster risk reduction.
02.
View allPublications
19 journal articles, reviews, and data descriptors, alongside book chapters, conference work, and preprints. Explore the research, supporting data, and open-source implementations.
BAM: A Physics-Informed Self-Supervised Framework for Near-Real-Time Wildfire Burned Area Mapping from Multi-Source Earth Observation
International Journal of Applied Earth Observation and Geoinformation
03.
About Me
I'm Mirza Waleed, a computational geographer who found early on that satellite pixels could carry the weight of saving lives, long before I had a word like "GeoAI" for it. That instinct became a career: building machine learning systems that turn Earth observation data into flood warnings and decision-ready tools.
My core project, GFSM, is a 30-metre global flood susceptibility map trained on tens of millions of samples — alongside NFSEA (near-real-time flood mapping), BAM (self-supervised burned area mapping), and PI-SwinIR (physics-informed terrain refinement). This work has taken me from Google Earth Engine to Shaheen III supercomputing and earned me recognition as a Google Developer Expert in Earth Engine (2023–2026), with findings published across disaster-risk, remote sensing, and environmental journals.
I hold a PhD in Geography (Remote Sensing & GeoAI) from Hong Kong Baptist University, and I'm still chasing the same question that started it all: how do we make climate risk visible before it becomes a disaster?
Tools & Technologies
Cloud & HPC Infrastructure
- Google Earth Engine
- Shaheen III Supercomputer
- Google Cloud / Vertex AI
- Docker
Machine Learning & AI
- XGBoost / LightGBM
- PyTorch
- Self-Supervised Learning
- Ensemble Methods
Remote Sensing & Earth Observation
- Multi-Sensor Fusion (SAR + Optical)
- Landsat / Sentinel / MODIS
- ERA5 Reanalysis Data
- Time-Series Analysis
Geospatial Data Engineering
- Rasterio / Xarray
- GeoPandas
- PostGIS / Spatial SQL
- Reproducible Pipelines
Programming Languages
- Python
- JavaScript
- R
- SQL
Geospatial Platforms
- QGIS
- ArcGIS Pro
- ENVI
- SNAP
Tools & Technologies
Cloud & HPC Infrastructure
- Google Earth Engine
- Shaheen III Supercomputer
- Google Cloud / Vertex AI
- Docker
Machine Learning & AI
- XGBoost / LightGBM
- PyTorch
- Self-Supervised Learning
- Ensemble Methods
Remote Sensing & Earth Observation
- Multi-Sensor Fusion (SAR + Optical)
- Landsat / Sentinel / MODIS
- ERA5 Reanalysis Data
- Time-Series Analysis
Geospatial Data Engineering
- Rasterio / Xarray
- GeoPandas
- PostGIS / Spatial SQL
- Reproducible Pipelines
Programming Languages
- Python
- JavaScript
- R
- SQL
Geospatial Platforms
- QGIS
- ArcGIS Pro
- ENVI
- SNAP
04.
Experience
Ph.D. in Geography (Remote Sensing & GeoAI)
Hong Kong Baptist UniversitySep 2022 - Aug 2026Hong Kong
- Completed doctoral research on scalable flood susceptibility modeling using ensemble machine learning and cloud-based geospatial computing.
- Developed the Global Flood Susceptibility Map (GFSM v1), a 30-meter resolution dataset recognized as a Web of Science Top 0.1% Highly Cited (Hot) Paper.
- Published multiple first-author papers in peer-reviewed journals including the International Journal of Disaster Risk Reduction and Environmental Impact Assessment Review.
- Served as Teaching Assistant for undergraduate courses in environmental hazards and health geography.
Google Developer Expert
Google2023 - PresentGlobal
- Recognized as one of approximately 50 Google Developer Experts worldwide for Google Earth Engine, based on contributions to the geospatial research and developer community.
- Delivered workshops and technical talks on cloud-based geospatial analysis, including tutorials on flood mapping, land surface temperature retrieval, and land cover classification using Earth Engine.
- Mentored researchers and developers on adopting Google Earth Engine for environmental monitoring and disaster risk assessment through open-source projects and educational content.
Current Position
Visiting Scholar
KAUST2024 - 2025Saudi Arabia
- Developed a self-supervised deep learning pipeline for high-resolution flood mapping, trained on the Shaheen III supercomputer — one of the top 10 most powerful systems globally.
- Investigated model-data fusion techniques, explainability, and uncertainty quantification to improve flood model transferability across diverse hydro-climatic regions.
- Collaborated with the Urban Analytics Lab on research integrating remote sensing, computational geography, and environmental risk assessment.
Freelance GIS Consultant
Fiverr2019 - 2025Remote
- Completed over 40 geospatial consulting projects with a consistent 5/5 client rating, covering remote sensing analysis, Earth Engine development, and spatial data processing.
- Delivered end-to-end solutions including satellite image classification, flood risk mapping, land cover change detection, and custom GEE application development for international clients.
- Translated complex geospatial analysis requirements into clear, reproducible workflows tailored to each client's domain and data constraints.
05.
View allPosts
awardsDec 1, 20252 min
Spatial Impact Award 2025
By Atlas.co, the Spatial Impact Award recognizes outstanding projects that leverage geospatial technology to address real-world challenges. In 2025, the award highlights innovative applications in environmental monitoring, disaster management, and sustainable development.
GEEPythonRemote Sensing
workshopSep 22, 20251 min
GIS Applications for Data-Driven Water Resource Management
A Hands-On Introduction with Flood Examples
GEEPythonRemote Sensing
awardsDec 1, 20242 min
Top GIS Voices 2024
Selected by Atlas.co as one of the 50 most influential voices in the GIS community. This award recognizes individuals who are actively sharing knowledge, fostering innovation, and driving the geospatial industry forward through community engagement and technical expertise.
GEEPythonGeoAI
06.
View allCartographic Work
Thematic maps and data visualizations produced for research publications, the 30 Day Map Challenge, and geospatial workshops.
Lines (1)Hexagons (1)Population (1)Africa (1)
Contact
07. Let's Collaborate
I am open to research collaborations, consulting, and speaking on GeoAI, flood risk and disaster intelligence, and cloud-native Earth observation — from academic partnerships to applied geospatial projects.
Email:waleedgeo@outlook.com