Automated Satellite Imagery Processing

Welcome to Vantage. This is an automated orbital monitoring laboratory.

Unlike traditional photography, these visuals are constructed from multi-spectral satellite data (ESA Sentinel-2) to highlight environmental shifts that are invisible to the naked eye. These are snapshot comparisons. Sampling is overlaid and a delta is created.

Water Vapor

Water vapor is fickle. It literally is a wisp in the wind, differing at times minute to minute. What this tooling does is it will take ~15 samplings of a median from a starting year. It will then request those data points samples through the API. The water vapor measurements will be ovelaid and a median of the year’s median will be created. This creates a baseline without having to send multiple requests to the CDSE API for large datasets that may not necessarily be adequate. It would be better if I could just snag everything they got and store their data myself

Without blasting them with requests for large dataset downloads, the sampling median is used to create a “baseline”. That baseline is then used to create a delta for the “future” date. The future date is treated much the same way, multiple datasets are evaluted with their metadata to be solid choices, the requests for the data is made for the targets, then a median is created and overlaid on the starting year’s measurements. Best I got for now.

Understanding the Data

#

• NDVI (Vegetation): High-precision math used to determine the density of chlorophyll.
• Impact Heatmaps: Autonomous detection of significant canopy loss.
• Atmospheric Masking: You may notice gray or white “redacted” areas in the imagery. These are pixels where clouds, smoke, or water vapor obscured the ground. To ensure data integrity, our pipeline automatically masks these areas to prevent weather from being mistaken for land-cover change.

This project is a bridge between orbital science and public visibility.

Got a location you want to check out? Drop human me a message!