Calculate total flooded area before and after heavy rains near Nanchang, Jiangxi, China by comparing supervised classified Landsat 8 imagery.
Google Slides presentation report featuring background/discussion, data acquisition and preprocessing steps, supervised classification analysis, statistical analysis and accuracy assessment.
Imagery analysis (supervised classification, ground truthing from samples, training data curation), Remote Sensing data acquisition & preprocessing (cloud cover masks), Statistical analysis, Cartography, GIS, ArcMap, research
Viewed several bands of multispectral imagery for the two images (before and after) to select training data for supervised classification, compared the difference in pixels classified as “Water” between the two images and then conducted an accuracy assessment for each classified image by sampling 100 random points. The number of “water” pixels increased by over 200% post-storm. “Wetlands” area decreased by almost 50%.
China Floods Presentation (pdf)
