Ongoing Projects

Used multi-temporal satellite data, InVEST, GIS, and CA-ANN modeling to assess and predict carbon storage, LST, and NDVI dynamics and their spatial interrelationships under rapid LULC change in Gazipur, highlighting impacts of urban expansion on ecosystem and thermal conditions.

The study aims to develop an automated framework for modeling the cascading impacts of landslide-dam formation and subsequent outburst floods. Pushing beyond traditional GIS, the methodology leverages Geospatial Artificial Intelligence to fuse multimodal Synthetic Aperture Radar and optical satellite imagery. This integration, utilizing deep learning techniques like convolutional neural networks, will enable automated dam detection and be coupled with hydraulic modelling to simulate flood propagation pathways. The expected outcome is a scalable, semi-automated system capable of supporting near-real-time hazard assessment and early warning, thereby improving risk management for cascading geohazards in vulnerable regions.

This research aims to develop an AI-integrated framework for rapid post-disaster assessment and community-led recovery mapping in coastal Bangladesh. The methodology applies Python and machine learning techniques to analyze satellite imagery, enabling swift impact analysis on both society and ecosystems. By fusing these automated geospatial analyses with participatory GIS methods that gather local knowledge, the project seeks to create a hybrid tool. The intended outcome is a practical system that accelerates damage assessment while ensuring community priorities guide recovery planning, thereby supporting more resilient and equitable post-disaster responses. 

This project aims to comprehensively assess the quality and suitability of groundwater for drinking and irrigation in a coastal region in Satkhira, Bangladesh, by analyzing its hydrogeochemical characteristics and identifying the major sources of contamination. Groundwater samples were collected from tubewells across six unions and analyzed for major ions, trace elements, and physicochemical parameters. Data were processed using Positive Matrix Factorization (PMF) to identify contamination sources and contributions, and Principal Component Analysis (PCA) was applied to interpret influencing factors. Suitability was evaluated against WHO and national standards using indices such as Total Hardness, Sodium Adsorption Ratio (SAR), Permeability Index (PI), Chloroalkaline Index (CAI), and Residual Sodium Carbonate (RSC). There is a high potential for widespread unsuitability of both drinking and irrigation, with significant salinity, elevated arsenic and iron levels, and very hard water characteristics.