This repo includes the code used in the Rio Grande do Sul 2024 Flood impact detection project.
This repo has the following objectives:
- to access and manage the data about the 2024 floods in Rio Grande do Sul
- to improve the quality of the available data sets by perfoming spatial and index-based intersections
- to repack the available information, thus providing actionable data for analysis
- (in the future) to implement analysis on the existing data (TBD)
Additionally, each notebook in the repo has its won set of goals
a. to access Sentinel 2 data.
b. to delinete the maximum flood area (MFA) from the 2024 flood events.
c. to import buildings and roads from OSM in the MFA.
d. to output the quantification of buildings and the list of roads per municipality in the MFA.
Roadmap
- Download Sentinel 2a data from https://browser.dataspace.copernicus.eu
In ArcGIS Pro (done!) - With de NDWI from Copernicus, explore bands and values to determine flood extent threshold (used 2 Natural Breaks, threshold = 0.662131)
- Reclassify the raster, to binary values (0, if < 0.55, 1, if >= 0.55)
- Run Majority Filter (8 neighbours, half rule)
- Polygonize the raster
- Remove cloudcover false positives, remove value = 0, remove area < 30 hectares
In Python: - Generate individual polygons for each affected municipality
- Download OSM roads and buildings for each affected municipality
- Identify OSM road tracts and buildings intersecting with flooded area (outer join)
a. to repack the Google Buildings data set available in https://zenodo.org/records/11243161 per municipality
b. to identify each building with the municipality it belongs to
c. to provide individual (i.e., per municipality) data sets from the original
a. to explore the geometric inconsistencies in the maximum flooded area composite cosntructed by Dr. Bárbara Giaccom
b. to correct the geometric errors
c. to dissolve the overlapping or touching polygons, simplifying the geometry and export it.
Feel free to reuse the code if needed, citing the original. Data licenses follow the original providers' ones.
Reach out to our team:
- alexandre.santos(at)lmu.de
- c.mirbach(at)lmu.de
Dr. Alexandre Pereira Santos Charlota Mirbach
Ongoing.