coastal-resilience-utilities 0.1.20

A utility package for coastal resilience analysis.

Coastal Resilience Utilities

This repository contains utilities for coastal resilience analysis and design.

Contents

• damage_assessment: damage assessments based on earth observation data and National Structure Inventory.
• damage_assessment_usvi: damage assessments based on low-res NSI grids + OpenBuildings
• mosaic: Mosaicking datasets.
• summary_stats: Summarize raster datasets using vector points and polygons.
• utils: A variety of functionality for doing geospatial analysis. Rasterize, vectorize, fetch features from OSM/OpenBuildings/ArcOnline, extract values from rasters.

Installation

The recommended way to install this repo is to use the provided Makefile to build a Docker image.

For example:

EXTRA_DOCKER_ARGS="-v <YOUR DATA DIRECTORY>:/data -v /cccr-lab:/cccr-lab" make bash-terminal

There are two env files:

• .env, which at this moment just contains GCP credentials
• .data.env, which maps files necessary for damage assessment

Note that it's a little tricky to guarantee out-of-the-box until we get more developers actively using this repo, but I do attempt to deploy this on multiple computers and make sure things work. If you have trouble let's work it out!

Examples

Check out the testing directory for examples on running analysis.

There are two other relevant repos in Gitlab:

• Datacube, which basically scales raster analyses
• GeospatialServices, which runs a variety of servers such as:
  • Jupyter for an analysis env
  • Prefect for reproducibility
  • Tile servers for app development

You don't need either to run, but the examples in GeospatialServices/Jupyter/notebooks are useful to understand using damage assessment.

Schematic

The Figjam schematic can be found here