Planetary-Scale Cloud Processing
Google Earth Engine provides access to a comprehensive planetary catalog of satellite imagery and Earth observation data with parallelized cloud computation. By leveraging GEE's cloud architecture, I execute temporal reductions and spectral transformations across multi-year Sentinel-2, Landsat (5, 8), and MODIS collections for remote sensing and satellite data analysis.
I utilize both the GEE JavaScript Code Editor for interactive prototyping and the Python GEE API (`ee` and `geemap`) for automated data extraction pipelines integrated with Jupyter and Python geospatial data processing using GeoPandas and Rasterio.
Demonstrated GEE Technical Capabilities
Time-Series Reductions & Composites
Writing custom reducer pipelines (median, percentile, min-cloud score) across ee.ImageCollection to generate radiometrically balanced, cloud-free seasonal mosaics across regional boundaries.
Cloud Machine Learning Classifiers
Deploying ee.Classifier.smileRandomForest and ee.Classifier.libsvm on satellite feature stacks for rapid regional land cover classification comparable with QGIS spatial analysis.
Environmental Index Calculation
Formulating automated expressions for NDVI, NDWI, MNDWI, NDBI, and EVI time-series charting to track drought, flood inundation extents, and vegetative vigor over historical windows.
Geemap & Python Integration
Connecting GEE cloud assets directly with local data science toolkits, automating bulk asset exports to Google Cloud Storage or Google Drive as Cloud-Optimized GeoTIFFs.
Multi-Temporal Satellite Classification & Urban Growth
Utilized multi-temporal satellite sensors to analyze 24 years of urban transformation in Ahmedabad across 5 epochs, validating land-cover changes against cloud-reduced composite tiles.