Md Khadem Ali

Geospatial Analysis of Urban Heat Island and Land Surface Characteristics in Mogadishu (2025)

This project presents a comprehensive geospatial analysis of the Urban Heat Island (UHI) effect in Mogadishu, Somalia, using satellite-based remote sensing and GIS techniques. Multi-source data derived from Landsat 8 Collection 2 Level-2 were processed in Google Earth Engine (JavaScript) to generate key environmental indicators, including Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), Land Surface Temperature (LST), and Urban Heat Island (UHI) intensity. An unsupervised land use/land cover (LULC) classification approach was applied using K-means clustering to identify major land surface types such as vegetation, built-up areas, bare land, and water bodies. The results reveal a strong spatial relationship between urban expansion and elevated surface temperatures, with high UHI intensity observed in densely built-up regions, while vegetated and water-covered areas exhibit significant cooling effects.

Research Conducted Under: Climate and Environmental Research Institute - CERI
Map Analysis and Cartography by: Md Khadem Ali
Data Source: Landsat 8 Surface Reflectance & Thermal (Collection 2 Level-2)
Tools Used: Google Earth Engine (JavaScript), ArcGIS Pro
Coordinate System: WGS 1984 / EPSG:4326

Mapping Surface Thermal Dynamics: A Land Surface Temperature Study of Pabna District (2013–2025)

This project presents a GIS-based Land Surface Temperature (LST) analysis of Pabna District, Bangladesh (2013–2025). Satellite-derived thermal data processed in Google Earth Engine (JavaScript) were used to generate year-wise LST maps, which were further analyzed and visualized using Python and ArcGIS Pro. The temperature values were classified into five categories: very low (<22°C), low (22–26°C), moderate (26–30°C), high (30–34°C), and very high (>34°C). The analysis indicates relatively lower temperatures in 2013 and 2015, while higher temperature zones became more prominent from 2019 to 2025, highlighting a temporal shift in surface thermal patterns across the study area.

Map Analysis and Cartography by: Md Khadem Ali
Data Source: Landsat 8/9 Surface Temperature (Collection 2 Level-2)
Tools Used: Google Earth Engine (JavaScript), Python, ArcGIS Pro
Coordinate System: WGS 1984 / EPSG:4326

Flood Risk Assessment Using ML in Chandai Union

I recently completed a Flood Risk Assessment for Chandai Union, Natore using Remote Sensing and Machine Learning. The project integrates Sentinel-2 NDWI, SRTM DEM (elevation & slope), and monsoon data from June to September 2023. A Weighted Overlay Model (NDWI × 0.5, Elevation × −0.3, Slope × −0.2) was applied along with a Random Forest Classifier (100 trees) to map flood-prone areas. The analysis achieved an overall accuracy of 98.35%, a Kappa coefficient of 0.967, and an AUC score of 0.998, with NDWI being the most important feature (268.99), followed by slope and elevation, confirming that low-lying, waterlogged areas carry the highest flood risk.

Data Analysis and Mapping by: Md Khadem Ali
Tools Used: Python (Pandas, NumPy, Matplotlib, Seaborn, GeoPandas), Google Earth Engine (JavaScript)

Precision Agriculture and Crop Yield Prediction

This project demonstrates how Artificial Intelligence and Machine Learning can support Precision Agriculture and Crop Yield Prediction. Using environmental variables like soil moisture, temperature, rainfall, humidity, and NDVI, I developed Random Forest models to accurately predict crop yields. The project integrates Python and MATLAB workflows, includes interactive visualizations, correlation analysis, feature importance, and exploratory data analysis. The workflow provides a research-ready framework for actionable insights in agriculture.

Data Analysis and Mapping by: Md Khadem Ali
Tools Used: Python (Pandas, NumPy, Matplotlib, Seaborn, GeoPandas), MATLAB

Geospatial Analysis of Air Quality in Bangladesh (2000–2025)

This project presents a comprehensive GIS-based analysis of air quality across Bangladesh from 2000 to 2025. Hourly AQI measurements along with key pollutants such as PM2.5, PM10, CO, NO₂, SO₂, and O₃ were integrated into spatial heatmaps and statistical visualizations to identify pollution hotspots, temporal trends, and correlations among pollutants.

Data Analysis and Mapping by: Md Khadem Ali
Data Source: National Air Quality Monitoring Dataset (2000–2025)
Tools Used: Python (Pandas, GeoPandas, Matplotlib, Seaborn)
Coordinate System: WGS 1984 / EPSG:4326

Geospatial Analysis of Environmental Factors and Drought Conditions in Chandai Union (2025)

This project presents a GIS-based environmental and drought assessment of Chandai Union (2025). Annual rainfall derived from CHIRPS data was classified to identify drought zones, while elevation, slope, NDVI, and land use/land cover (LULC) layers were integrated to analyze environmental variability and spatial patterns. The study highlights the relationship between rainfall distribution, vegetation health, terrain characteristics, and drought vulnerability within the study area.

Map Analysis and Cartography by: Md Khadem Ali
Data Source: CHIRPS Rainfall, SRTM DEM, Sentinel-2 NDVI, ESA WorldCover
Tools Used: Google Earth Engine, ArcGIS Pro
Coordinate System: WGS 1984 / EPSG:4326

Spatial Variation of Land Surface Temperature in Pabna District: 2010–2025

This project examines the spatial and temporal variations of land surface temperature (LST) in Pabna District from 2010 to 2025. MODIS Terra LST data were processed in Google Earth Engine to produce annual mean rasters, while NDVI was derived to assess vegetation influence on surface temperature. The analysis highlights a clear warming trend in urban and sparsely vegetated areas, whereas regions with dense vegetation experienced comparatively smaller temperature increases, revealing the impact of land cover on local thermal dynamics.

Map Analysis and Cartography by: Md Khadem Ali
Data Source: MODIS Terra LST (MOD11A2) & NDVI (MOD13A2)
Tools Used: Google Earth Engine, ArcGIS Pro
Coordinate System: WGS 1984 / EPSG:4326

Spatiotemporal Modeling of Flood-Induced Crop Vulnerability: A Geo-Digital Twin Approach
(Case Study: Natore District, Bangladesh)

This project represents my independent work on assessing flood-induced vulnerability of Aman and Boro rice crops in Natore District, Bangladesh. I developed a Geo-Digital Twin framework that combines synthetic environmental modeling (elevation, river influence, flood depth) with statistical crop damage simulations to quantify risk levels across the district. The analysis revealed that Aman rice exhibits an average yield loss of 47.85% (Std. Dev: 11.91%, Min: 4.55%, Max: 95.31%), while Boro rice shows an average loss of 40.60% (Std. Dev: 11.38%, Min: 0.95%, Max: 83.97%). The workflow includes generating a high-resolution spatial grid, modeling river influence on flood dynamics, simulating topographic elevation, estimating flood depth, and predicting yield losses for both Aman and Boro rice. Risk classification indicated that 42.01% of Aman rice areas and 20.25% of Boro rice areas fall under High Risk, while 57.22% of Aman and 76.49% of Boro rice areas are under Moderate Risk, and only a small fraction (0.77% for Aman, 3.26% for Boro) are at Low Risk. Correlation analysis showed that flood depth (Aman: 43.7%, Boro: 40.6%) and river influence (Aman: 52.7%, Boro: 48.6%) positively drive crop damage, while elevation is negatively correlated (Aman: -51.1%, Boro: -47.4%). Spatiotemporal visualizations, including maps, heatmaps, and proportional risk charts, were independently produced using Python libraries (GeoPandas, Matplotlib, Seaborn, Pandas, NumPy). This project demonstrates my ability to integrate geospatial analysis, statistical modeling, and data visualization into a coherent, research-grade framework for actionable flood risk assessment in agriculture.

Data Source: GADM shapefiles for Natore District, synthetic data for flood and crop modeling
Tools Used: Python (GeoPandas, Matplotlib, Seaborn, Pandas, NumPy)

Crop Health Assessment of Rajshahi Division (2025)

This project evaluates crop health conditions across Rajshahi Division for the year 2025 using satellite-based vegetation indices. Sentinel-2 surface reflectance imagery was processed to derive NDVI for assessing vegetation vigor and the Vegetation Condition Index (VCI) for identifying crop stress related to moisture deficiency. The results reveal distinct spatial patterns of healthy and stressed croplands, providing valuable insights for agricultural planning and early warning of drought-prone areas.

Data Source: Sentinel-2 SR (ESA)
Tools: Google Earth Engine, ArcGIS Pro

Spatio-Temporal Analysis of Earthquakes in Bangladesh (1826–2025)

The "Spatio-Temporal Analysis of Earthquakes in Bangladesh (1826–2025)" project, conducted by Md Khadem Ali, provides a comprehensive 200-year overview of seismic activity across Bangladesh. Using USGS earthquake records and geospatial visualization techniques, this project maps all recorded earthquakes, showing both magnitude and spatial distribution. The interactive map and heatmaps highlight high-density seismic zones, while decade-wise animations and magnitude-frequency analyses facilitate research-driven insights into temporal trends and risk assessment. This visualization supports disaster preparedness, research, and educational purposes, with clear markers, color-coded magnitude scales, and an intuitive layout for professional presentation.

Data Source: USGS Earthquake Catalog (1826–2025)
Tools Used: Jupyter Notebook, Colab

Urban Expansion over Agricultural Land in Pabna District (2015–2025)

This project analyzes the conversion of agricultural land into built-up areas in Pabna District between 2015 and 2025 using satellite-based change detection techniques. Landsat 8 and Landsat 9 imagery were processed to compute the Normalized Difference Built-up Index (NDBI), enabling the identification of urban expansion patterns. The results show a noticeable increase in built-up areas, particularly around urban centers and major transportation corridors, highlighting the pressure of unplanned urbanization on agricultural land.

Data Source: Landsat 8 & Landsat 9 (USGS)
Tools: Google Earth Engine, ArcGIS Pro

VIIRS DNB Night-Time Light Radiance of Bangladesh (2024–2025)

This project maps annual night-time light pollution across Bangladesh using the VIIRS DNB monthly radiance dataset for the period January 2024 – December 2025. The study highlights urban and rural light intensity differences and identifies major urban hubs like Dhaka and Chattogram. This analysis supports applications in urban monitoring, environmental planning, and assessing the spatial extent of artificial lighting.

Data Source: VIIRS DNB (NASA/NOAA)
Tools: Google Earth Engine, ArcGIS Pro

Winter Vegetation Change in Rajshahi Division (December 2024–2025)

This project analyzes winter vegetation change in Rajshahi Division, Bangladesh using Sentinel-2 Surface Reflectance data. Conducted by me, Md Khadem Ali, the study computes NDVI for December 2024 and December 2025, applies cloud masking with the Scene Classification Layer (SCL), and generates median composites in Google Earth Engine (GEE). Vegetation dynamics were quantified using ΔNDVI (NDVI₍2025₎ − NDVI₍2024₎), highlighting areas of vegetation gain and loss. The resulting GeoTIFFs were visualized and analyzed in ArcGIS Pro using a diverging color palette to clearly depict regions of decline (red), stability (white), and increase (green/blue).

Data Source: Sentinel-2 SR (NASA / ESA)
Tools: Google Earth Engine, ArcGIS Pro

Seasonal Disruption in Natore District (2001–2024)

Based on this project, a research paper has been published
DOI: 10.2139/ssrn.6055754

This project, Seasonal Disruption Mapping of Natore District, was conducted by me, Md Khadem Ali, using Google Earth Engine (GEE) and ArcGIS Pro. The study analyzes long-term vegetation seasonality changes from 2001 to 2024 using MODIS-derived NDVI data. Seasonal vegetation dynamics were examined separately for winter and monsoon periods and combined to assess the degree of seasonal rhythm disruption across the district.

The final output classifies Natore into five disruption levels, Very High, High, Moderate, Low, and Stable, highlighting areas where natural seasonal patterns have weakened significantly. Supporting maps of Winter NDVI, Monsoon NDVI, and NDVI Difference (Monsoon − Winter) were used to validate spatial consistency. This analysis supports applications in climate impact assessment, agricultural monitoring, and regional environmental planning.

Data Source: MODIS NDVI (NASA)
Tools: Google Earth Engine, ArcGIS Pro

Seismic Risk Map of Bangladesh (2001-2025)

The project titled "Seismic Risk Map of Bangladesh (2001-2025)" was created by me, Md Khadem Ali, using ArcGIS Pro. This project visualizes the earthquake vulnerability of Bangladesh over the last 25 years, highlighting areas ranging from very low to very high seismic risk. By presenting earthquake data in a clear and structured way, the map provides valuable insights into the spatial distribution of seismic activity across the country.

Data Source: USGS Earthquake Catalog | Software: ArcGIS Pro

Flood Risk Analysis of Pakistan (2022–2025)

This project assesses flood risk across Pakistan for the years 2022 and 2025 using a multi-criteria geospatial modeling approach. Annual precipitation, elevation, and distance to river networks were integrated to generate a normalized flood risk index and classify areas into low, medium, and high flood risk zones. The results support flood hazard assessment, disaster risk reduction, and climate impact studies.

Data Sources: CHIRPS, Copernicus DEM, HydroSHEDS
Tools: Google Earth Engine

The Rivers of the Queen's Land

The project titled "The Rivers of the Queen's Land" was conducted by me, Md Khadem Ali, using ArcGIS Pro. This project maps and visualizes the river networks of Natore District, highlighting the intricate waterways that have historically shaped the region's landscape. By analyzing and presenting the rivers through spatial visualization and hydrology-focused mapping, the project provides insights into the interaction between natural waterways and human settlements. The visualization emphasizes Natore's unique geography, its wetlands, and the historical significance of rivers in connecting communities and supporting livelihoods.

Data Source: HydroSHEDS | Software: ArcGIS Pro

Bouguer Gravity Anomaly of Bangladesh

The project titled "Bouguer Gravity Anomaly of Bangladesh" was conducted by me, Md Khadem Ali, using ArcGIS Pro. This project converts point-based gravity measurements into raster surfaces, generates contour lines, and visualizes subsurface density variations using a five-class color scheme. This approach provides a clear and interpretable view of gravity anomalies across Bangladesh, highlighting variations in subsurface structures and tectonic features. Designed with a professional color palette and clear legends, this visualization supports applications in geophysical interpretation, tectonic analysis, and spatial decision-making.

Data Source: U.S. Geological Survey (USGS) | Software: ArcGIS Pro

Baraigram Upazila LULC Analysis 2024

The project titled "Baraigram Upazila Land Use / Land Cover (LULC) Analysis 2024" was conducted by me, Md Khadem Ali, using ArcGIS Pro. This project classifies the area into seven land cover classes: Water, Trees, Flooded Vegetation, Crops, Built Area, Bare Ground, and Rangeland, based on Sentinel-2 imagery. To enhance readability and interpretability, histogram plots were created using Python (Matplotlib), displaying the counts for all classes. This approach ensures even smaller classes, such as Rangeland and Flooded Vegetation, are clearly visualized. Designed with a professional color palette and clear legends, this visualization provides a comprehensive overview of the land cover distribution in Baraigram Upazila, supporting applications in environmental planning, agricultural assessment, and geospatial analysis.

Data Source: Sentinel-2 imagery | Software: ArcGIS Pro & Python (Matplotlib)

South Carolina LULC Change Prediction (2019–2030)

The project titled "South Carolina Land Use & Land Cover Change Prediction (2019–2030)" was conducted by me, Md Khadem Ali, using a combination of Google Earth Engine and geospatial analysis techniques. This project visualizes the existing land cover of South Carolina for 2019 and predicts possible changes by 2030 through a rule-based urban growth model. The 2019 map highlights baseline land cover categories such as Water, Urban, Forest, Agriculture, Grassland, Shrubland, Barren land, and Wetlands, while the 2030 predicted map shows potential urban expansion and landscape changes over the next decade. Integrating spatial maps with predictive modeling provides an intuitive understanding of land cover dynamics, making it suitable for research, planning, and decision-making purposes. The visualizations are designed with clear legends, distinctive color coding, and a professional layout to enhance readability and interpretation.

Data Source: USGS NLCD 2019, TIGER/US Census Boundaries, Google Earth Engine

Bangladesh Topography & Water Map

The project titled "Bangladesh Topography & Water Map" was conducted by me, Md Khadem Ali, using Google Earth Engine (GEE). This project visualizes the elevation patterns and hydrological landscape of Bangladesh through high-resolution digital elevation data and river network overlays. It integrates COPERNICUS DEM (GLO-30)/ for topographic elevation, WWF HydroSHEDS for rivers and drainage systems, and USDOS LSIB (2017) for national boundaries. The map classifies the country into six elevation-based zones, Floodplain / Lowland (<5 m), Agricultural Plains (5–10 m), Upland / Medium (10–30 m), Hills (30–60 m), Hilltops (>60 m), and Rivers & Water Bodies, offering a comprehensive view of Bangladesh's physiographic diversity from the coastal deltaic plains to the eastern hills. Designed with a professional color palette and clear legends, this visualization supports applications in environmental planning, agricultural assessment, and climate resilience research.

Data Source: COPERNICUS DEM (ESA), WWF HydroSHEDS, USDOS LSIB (2017)

Columbia, SC Precipitation Analysis (2015–2024)

The project titled "Columbia, SC Precipitation Analysis (2015–2024)" was conducted by me, Md Khadem Ali, using a combination of Python (GeoPandas, Pandas, Matplotlib, Contextily, Meteostat) and Google Colab. This project visualizes the annual precipitation trends of Columbia, South Carolina over the past decade through a detailed city map and a complementary bar chart for year-wise rainfall comparison. The map highlights the location of Columbia within South Carolina, while the bar chart provides a clear quantitative summary of total precipitation in millimeters and inches for each year. Integrating spatial maps with analytical charts allows for an intuitive understanding of rainfall patterns, making it suitable for educational purposes, presentations, and analytical insights. The visualizations are designed with clean legends, gradient color coding, and footers to maintain a professional and readable layout.

Data Source: Meteostat (2015–2024) & PublicaMundi / Natural Earth GeoJSON

World Continent Population Visualization (2024)

The project titled "World Continent Population Visualization (2024)" was conducted by me, Md Khadem Ali, using a combination of Python (GeoPandas, Pandas, Matplotlib) and Google Colab. This project visualizes the population distribution of continents globally through a world map and complements it with a horizontal bar chart for a continent-wise comparison. The world map highlights population differences across continents, while the bar chart provides a clear quantitative summary of total population in millions for each continent. Integrating spatial maps with analytical charts allows for an intuitive understanding of global population patterns, making it suitable for educational purposes, presentations, and analytical insights. The visualizations are designed with clean legends, gradient color coding, and footers to maintain a professional and readable layout.

Data Source: World Bank (2024) & Natural Earth (1:110m Cultural Vectors, 2024)

Population Density Analysis of Rajshahi Division (2020)

The project titled "Population Density Analysis of Rajshahi Division" was conducted by me, Md Khadem Ali, using a combination of Google Earth Engine (GEE) and Python (matplotlib / Google Colab). This project visualizes both the spatial distribution of population across the division and a district-wise comparison through analytical bar charts. The GEE map highlights high-density urban areas such as Sirajganj, Pabna, and Rajshahi City, while rural districts like Chapai Nawabganj and Naogaon exhibit relatively lower population densities. Complementing this, the Python-based bar chart provides a clear quantitative comparison, enabling quick identification of population patterns. This visualization demonstrates how integrating spatial maps with statistical charts can yield actionable insights for urban planning, resource allocation, and policy-making.

Data Source: GPWv4 (CIESIN, 2020)

LULC, NDWI & Water Bodies of Natore District

Project ID: SP-7/24

The map titled "LULC, NDWI & Water Bodies of Natore District" was prepared by me, Md Khadem Ali, as part of the project "Space Based Observation of Wetlands of Bangladesh: Wetland Inventory, Assessment and Monitoring", implemented under the SERC (Space and Environment Research Center) Wetland Conservation Program. This map presents spatial analysis of Land Use Land Cover (LULC), Normalized Difference Water Index (NDWI), and extracted water bodies of Natore District, based on satellite imagery acquired on 27 March 2025. The analysis was performed using Landsat-8 data and spatial processing was conducted using the WGS 1984 coordinate system. All data processing, analysis, and map preparation were done using ArcGIS Pro. This project was supervised by Mithun Kumar, Scientific Officer & Head, Aeronautics & Space Applications Division, Space and Environment Research Center (SERC), Bangladesh.

Flood Risk Mapping

This project generates a Flood Risk Map for Natore District, Bangladesh using Google Earth Engine (GEE). It integrates Sentinel-1 SAR data, SRTM elevation, and Sentinel-2 NDVI to assess flood vulnerability. Risk zones are calculated by combining water presence, low elevation, and vegetation cover into a weighted index. The final map displays flood risk levels with green (low), yellow (medium), and red (high) zones for easy interpretation. This tool supports disaster preparedness and localized flood risk assessment.

Urbanization and Infrastructure Analysis

This project analyzes urbanization trends and infrastructure patterns in Natore District, Bangladesh, using satellite-based geospatial data. By integrating population density (CIESIN), land cover (MODIS), and administrative boundaries (FAO GAUL), it highlights urban growth and densely populated areas. The study supports urban planning by identifying zones requiring infrastructure development. Visual outputs include color-coded maps for population and urban areas, aiding in clear, data-driven decision-making.

Natore Zilla NDVI Analysis

This project analyzes vegetation health in Natore Zilla, Bangladesh using NDVI derived from Sentinel-2 imagery via Google Earth Engine. It provides full regional coverage with a custom color-coded NDVI map, highlighting variations in plant density and health. The workflow is scalable and offers exportable statistical outputs for further research. This analysis supports informed decision-making in agriculture and environmental sustainability.

Khulna-Sundarbans NDVI Analysis

This project analyzes the vegetation health of the Khulna-Sundarbans region using NDVI derived from Landsat 8 satellite imagery via Google Earth Engine. Focusing on the year 2022, it highlights vegetation density, forest coverage, and areas under environmental stress. The NDVI values are visualized with a custom color palette to distinguish healthy forests from degraded or barren areas. This analysis supports ecological monitoring and conservation planning in the Sundarbans.

GeoKhadem: A Lightweight Python Library for GIS & Remote Sensing

This is a private project. Code not publicly available.

GeoKhadem is a Python library developed by me to support basic GIS and Remote Sensing tasks such as NDVI/NDWI computation, raster analysis, and spatial data handling. It was designed as a personal challenge to build a reusable, modular geospatial toolkit for educational and research purposes. The project enhanced my understanding of package structure, documentation writing, and Python-based geospatial processing.

Developed by Md Khadem Ali

GeoLocation Analyzer (Chrome Extension)

I have developed a Chrome Extension that detects geographic location data such as place names and coordinates from any webpage, allowing users to instantly visualize them on an interactive map. The extension features right-click context menu integration, regex-based coordinate detection, and map visualization using Leaflet.js. This tool helps researchers, journalists, and GIS professionals quickly extract and analyze spatial data directly from web content.

Developed by Md Khadem Ali