Interactive Data Features on Indigenous Communities and Disaster Resilience and Recovery

Analysis supporting two interactive features relating to Indigenous communities and disaster recovery and resilience.

• Feature 1: https://www.urban.org/data-tools/disaster-data-federal-funding-strategies-tribes-indigenous-communities
• Feature 2: https://www.urban.org/stories/indigenous-communities-tribes-disaster-risks-recovery-resilience

Repository layout

The overarching project convention is that specific tasks are in dedicated .R files, which are then sourced in _feature1.qmd and _feature2.qmd, respectively. It is these .qmd files that generate the final data, charts, and statistics reflected in the two interactive features on urban.org. (Note that charts are included merely as references; actual interactive charts are developed separately, outside of this repository.)

Scripts, data, and metadata are organized as below:

scripts/
  feature_1/    # Interactive data tool featuring tribes' histories of presidentially-declared disasters
  feature_2/    # The charts and statistics included in the narrative feature
  utilities/    # shared helpers (name normalization, substring matching, coverage scoring)
data/
  nri-tracts/   # National Risk Index tract-level hazard data, cached locally for convenience
  crosswalks/   # For translating from source data geographies to our target geographies (primarily federally-recognized tribes)
outputs/        # Rendered figures and data reflecting what appears in the two features
renv.lock       # A snapshot of the specific package versions used to generate all outputs

How to run

Dependencies are pinned with renv. From an R session at the repo root:

renv::restore()

This will ensure that you have loaded all of the same package versions as were used to develop these data originally.

Feature 1

The primary work of Feature 1 is to relate FEMA presidential disaster declarations data back to a consistent set of Indigenous entities. While this seems a simple task at the outset, in practice it entails dealing with inconsistently named and identified FEMA data, Indigenous entities, and Indigenous geographies.

Interested users can step through individual scripts in /feature_1 to better understand how exactly this process was implemented, but /feature_1/_feature1.qmd compiles the final set of Indigenous disaster declarations data that are likely to be of primary interest.

Feature 2

This feature is focused on assembling data describing Indigenous communities' prior disaster costs, disaster exposure and risk, and disaster assistance funding. Component scripts within /feature_2 handle data preparation for each figure--or, where a single dataset is used across multiple figures, for each dataset--and /feature_2/_feature2.qmd then sources these data preparation scripts to assemble and plot the final data and generate any statistics that are used in the feature. Note that the data for Feature 2 reliese on the universe of Indigenous entities created as part of the Feature 1 workflow, which should be run first.

At a high level, Feature 2 figures include:

Figure	Topic
1	SHELDUS cumulative hazard damages
2	Distribution of major disaster declarations
3	Per-capita estimated annual losses (NRI)
4	Distribution of community risk scores (NRI)
5	Awarded recovery funding (IHP + PA) by program
6	SBA approved loan amounts
7	Awarded HMA (mitigation) funding by program