On November 5, 2015 an iron ore tailings dam failed in Mariana, Minas Gerais, Brazil. The catastrophic flooding that resulted downstream resulted in the deaths of 19 people and severe environmental devastation.
Chenglin Liu has created a web map using ArcGIS API for JavaScript that shows changes in surface reflectance before and after the disaster. Data are from the Landsat 8 sensor and the images show false-color composites from dates immediately before and after the disaster.
False-color composites are images that assign reflectance values to the (screen-visible) Red, Green, and Blue that come from areas of the spectrum that are NOT Red, Green, and Blue. Here, Red on screen does show red band reflectance, but Green shows near-infrared reflectance, and Blue shows short-wave infrared reflectance.
Photosynthetically productive vegetation strongly reflects near-infrared reflectance, so the bright green areas here show high vegetation density. Brown shows fallow fields, and purple shows bare soil or built-up areas.
You can find the interactive map HERE. Clicking the "rainbow" layer button in the top right triggers a sliding feature in the map that allows you to compare Landsat images from just before, and just after the collapse. It also opens up a panel where you can toggle on/off a red feature that shows the affected area immediately downstream.
A paper by Aires et al. in the journal Land Use Policy provides a detailed description of the effects experienced downstream from the dam collapse and use perform multi-criteria evaluation in Idrisi software (developed at Clark University) to investigate potential local sites for the disposal of tailings.
תגובות