GIS, Remote Sensing and Geospatial Deep Learning
Remote sensing researcher at PUST, Bangladesh. I map trees, floods and cyclones with drones, satellites and deep learning.
Characterization of Urban Trees with Deep Learning Techniques in Pabna Municipality
Submitted to a peer-reviewed journal
Spatio-temporal Patterns of Lightning Events and Associated Casualties in Bangladesh: Linkages with Land Use and Land Cover
Submitted to a peer-reviewed journal
Impact of Armed Conflict on Ambient Air Quality: A Multi-Pollutant Spatiotemporal Analysis Across Iranian Provinces, 2022 to 2026
International Conference on Emerging Frontiers in Advanced Sciences and Technologies (EFAST 2026), Pabna, Bangladesh
Maps early warning last-mile gaps in the Jamuna-Padma char corridor: where floods, failed roads and submerged mobile towers cut settlements off at the same time. Built for the RIMES Mapathon 2026.
Python, Google Earth Engine, Sentinel-1, QGIS
Maps annual 10 m wind over Bangladesh from ERA5 reanalysis, with streamlines and wind roses for major cities.
Python, ERA5, Open-Meteo
Reconstructs 12 major cyclone tracks toward Bangladesh from 1970 to 2024 and turns them into a severity-weighted exposure corridor.
Python, geopandas, scipy, matplotlib
Code for my thesis: a canopy model trained on UAV imagery labels a city-wide satellite classifier, then crowns are converted to biomass, carbon and ward statistics for Pabna Municipality.
Python, PyTorch, Random Forest, UAV
2025 to present
Geospatial Lab of Environment and Disaster, Pabna University of Science and Technology. Supervised by D. Eng. Md Rahedul Islam, Professor.
2025
In collaboration with the JAXA Earth Observation Research Center and the Institute of Industrial Science, University of Tokyo.
2026
Pabna University of Science and Technology. CGPA 3.56 / 4.00 across seven semesters; coursework completed, final result awaited.
Open to research collaboration, mapping work and questions about any project.
