TraCR is developed by the National Institute of Standards and Technology (NIST) Community Resilience Program.
TraCR is a county-level database designed to support the development, testing, and tracking of indicators related to community resilience.
NIST reports that TraCR contains data for 3,230 counties and county equivalents across the United States and U.S. territories.
Data Publication
Tracking Community Resilience (TraCR) Database
Maria Dillard, Jarrod Loerzel, Donghwan Gu, Tiffany Cousins, Tzong Hao Chen
Contact: Tiffany Cousins
Identifier: <https://doi.org/10.18434/mds2-3978>
Described in these articles:
<https://ascelibrary.org/doi/full/10.1061/NHREFO.NHENG-1642>,
<https://ascelibrary.org/doi/10.1061/NHREFO.NHENG-2224>
Version: 1.0
First Released: 2025-11-18
Revised: 2025-11-18
Abstract
Community resilience is the ability to
prepare, adapt, withstand, and recover from disruptions.
There is a growing field of research that focuses on
community-level resilience.
Many have developed and studied methods for measuring resilience
quantitatively and qualitatively.
The Tracking Community Resilience (TraCR) database is a tool
for developing and testing analytical methods
for computing county-level indicators for community resilience.
TraCR can be used by local communities, researchers, and
decision makers at various levels to assess and
measure longitudinal community resilience.
TraCR is part of the Community Resilience Assessment Methodology,
which aims to develop a first-generation methodology
to assess resilience at the community scale based
on community functions, supported by buildings
and infrastructure systems, and
the recovery of those functions following a disruptive hazard event.
https://data.nist.gov/rmm/records?keyword=TraCR
https://data.nist.gov/od/ds/mds2-2297/NIST_Resilience_Indicator_Inventory_v.01.xlsx
https://data.nist.gov/od/ds/mds2-2297/NIST_Resilience_Indicator_Inventory_v.01_Data_Dictionary.xlsx
https://data.nist.gov/pdr/bulkdownload/
uv run python tools/pdrdownload.py -I mds2-3978
uv run python tools/pdrdownload.py -I mds2-3978 -DNIST Community Resilience Products
NIST Community Resilience Assessment Methodology
NIST Risk Reduction and Recovery Program
NIST distributes public research datasets through the NIST Public Data Repository (PDR).
The PDR provides persistent dataset records, metadata, downloadable data files, supporting documentation, and citation information.
TraCR is designed for county-level analysis.
NIST reports coverage for 3,230 counties and county equivalents, including locations in:
- the contiguous United States
- Alaska
- Hawaii
- Puerto Rico
- the U.S. Virgin Islands
Geographic identifiers should be preserved during processing so TraCR records can later be connected to other public datasets using standard geographic identifiers where available.
TraCR is intended to support analysis of community resilience indicators over time.
Available years vary by measure and underlying source dataset.
Missing years should not automatically be interpreted as zero values. Missingness should be preserved and evaluated during data processing and analysis.
TraCR contains measures used to develop and evaluate community resilience indicators.
The measures draw from multiple public data sources and represent aspects of community resilience across social, economic, physical, and related systems.
This project does not assume that all measures have identical units, time coverage, completeness, or interpretation.
Indicator and measure metadata should be retained whenever possible.
The processing layer may perform operations such as:
- standardizing column names
- preserving and standardizing geographic identifiers
- converting values to appropriate data types
- identifying time fields
- reshaping source data when needed for analysis
- retaining missing values
- preserving useful indicator and source metadata
Processed data are derived artifacts. They do not replace or modify the original raw source files.
The authoritative source for TraCR data and metadata is NIST.
This repository may cache copies of public source data in data/raw/
for reproducibility and educational use, but those copies are not
authoritative NIST publications.
When data files are downloaded, the project should retain enough provenance information to identify:
- source organization
- dataset name
- source URL
- download date
- source version or release, when available
- original filename
- applicable citation or DOI, when available
NIST data are U.S. government research data.
Users should consult the metadata, citation information, and terms provided with the specific TraCR release in the NIST Public Data Repository before redistributing or publishing derived datasets.
Project software in this repository is licensed separately under the MIT License.
Dillard, Maria , Loerzel, Jarrod , Gu, Donghwan ,
Cousins, Tiffany , Chen, Tzong Hao (2025),
Tracking Community Resilience (TraCR) Database,
National Institute of Standards and Technology,
https://doi.org/10.18434/mds2-3978
(Version: 1.0, Accessed 2026-09-02)
When using this software, see:
The repository currently uses the following TraCR source files:
data/raw/
├── TraCR_v1_database.csv
├── TraCR_Metadata.xlsx
├── TraCR_DataCoverageMatrix.xlsx
├── TraCR_ TechnicalSupportDocument.pdf
└── README.txt
TraCR_v1_database.csv is the primary analytical data file.
The database is stored in wide format. Each row contains a geographic identifier and year, followed by columns representing TraCR measures and indicators.
Examples include:
UID
fips
period
INFRA110007
INFRA120001
ECNVIT510001
NATENV410002
...
The project adapter reshapes these wide indicator columns into the canonical long-format schema used by downstream processing, analysis, views, and renderers:
geography_id
geography_name
indicator_id
indicator_name
unit
year
value
In this transformation:
fips -> geography_id
period -> year
indicator column name -> indicator_id
indicator cell value -> value
TraCR_Metadata.xlsx contains four worksheets, including:
- Measures - metadata for TraCR measures and indicators
- Column Metadata - descriptions, units, types, and other metadata for columns in the TraCR database
- Source Definitions - definitions of abbreviated source names and general source URLs
- an introductory worksheet describing the workbook and release
The Column Metadata worksheet provides the metadata currently used by this project to associate TraCR indicator identifiers with human-readable descriptions and units.
For example:
INFRA110007
% of households with broadband internet service
Because browser-based Marimo WASM applications should not depend on reading Excel workbooks at runtime, the required worksheet is exported to:
data/processed/TraCR_Metadata_Column_Metadata.csv
and copied for browser execution to:
src/public/TraCR_Metadata_Column_Metadata.csv
The original Excel workbook remains the source metadata artifact. The CSV is a browser-compatible derivative used by the application.
The Marimo WASM application requires browser-accessible copies of the runtime data files under:
src/public/
Current runtime data include:
src/public/
├── TraCR_v1_database.csv
└── TraCR_Metadata_Column_Metadata.csv
These files allow the deployed application to operate entirely in the browser without a Python server or database service.
Generated deployment artifacts are not authoritative source data.
The original NIST files retained under data/raw/
remain the source artifacts for reproducibility.
The TraCR database uses the fips field as its geographic identifier.
FIPS values must be treated as strings rather than integers so leading zeros
are preserved. For example:
01001
must remain 01001, not 1001.
The current TraCR database contains geographic identifiers but does not provide a complete human-readable geography-name lookup in the analytical CSV itself. The project therefore preserves the original FIPS identifier while a separate authoritative FIPS-to-geography-name lookup is incorporated.
Until such a lookup is available, the application may display geographic labels using the FIPS identifier itself rather than inventing geographic names.
A future geography lookup should be derived from an authoritative geographic reference source and stored as a small reproducible lookup table rather than being manually constructed.
To add human-readable county names, use the U.S. Census Bureau's official 2020 FIPS reference file.
-
Open the Census 2020 FIPS page:
-
Download:
2020 State, County, Minor Civil Division, and Incorporated Place FIPS Codes -
Save the downloaded file in:
data/raw/
The project can then generate a small browser-safe lookup such as:
data/processed/TraCR_Geography_Lookup.csv
with fields such as:
geography_id
geography_name
The lookup should be generated from the downloaded Census source rather than constructed manually.
This also avoids depending on Census network availability during application execution or build steps. The earlier direct-download attempt failed during DNS resolution before any Census data were retrieved.