Multispectral & Temporal Earth Observation
Remote Sensing is the science of acquiring qualitative and quantitative information about the Earth's surface without physical contact. Through systematic acquisition of electromagnetic reflectance data across Visible, Near-Infrared (NIR), Shortwave Infrared (SWIR), and Thermal bands, I analyze complex environmental and anthropomorphic transformations using planetary cloud platforms like Google Earth Engine remote sensing analysis.
My practice spans radiometric calibration, atmospheric correction (DOS1, Sen2Cor), cloud masking with QA bands, spectral index calculation, and supervised machine learning classification algorithms executed in QGIS spatial analysis and scripted via Python geospatial data processing.
Spectral Processing & Analytical Methodologies
Spectral Indices & Ratio Modeling
Computation of Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Normalized Difference Built-up Index (NDBI), and Enhanced Vegetation Index (EVI) for biomass and urban density tracking.
Multi-Temporal LULC Classification
Supervised land-use classification using training ROI samples, spectral signature separation, confusion matrix generation, and Kappa coefficient statistical validation across multi-year intervals.
Cloud Masking & Mosaic Synthesis
Filtering cloud shadow and atmospheric haze using pixel reliability QA bands to produce seamless, cloud-free seasonal composite tiles across extensive geographical study areas.
3D Surfaces & Elevation Integration
Combining multispectral satellite layers with LiDAR point cloud and 3D surface modeling to generate orthorectified elevation-corrected reflectance datasets.
Ahmedabad Multi-Decadal LULC Change Detection (2000–2024)
Conducted classification across 5 epochs (2000, 2008, 2015, 2020, 2024) categorizing Water, Vegetation, Open Land, Built-up, and Agriculture classes with rigorous cross-validation and transition matrix calculations.