Skip to main content
Dryad

Comparing first street foundation and PRIMo flood hazard data across the Los Angeles metropolitan region

Cite this dataset

Schubert, Jochen; Sanders, Brett; Mach, Katharine (2024). Comparing first street foundation and PRIMo flood hazard data across the Los Angeles metropolitan region [Dataset]. Dryad. https://doi.org/10.5061/dryad.kd51c5bcz

Abstract

Extreme flooding events are becoming more frequent and costly, and impacts have been concentrated in cities where exposure and vulnerability are both heightened. To manage risks, governments, the private sector, and households now rely on flood hazard data from national-scale models that lack  accuracy in urban areas due to unresolved drainage processes and infrastructure. The data in this repository supports an assessment of the uncertainties of First Street Foundation (FSF) flood hazard data, available across the U.S.. For the analysis, FSF data was compared to PRIMo-Drain, a flood hazard model that resolves drainage infrastructure and fine resolution drainage dynamics.

In the linked journal manuscript, using the case of Los Angeles, California, we find that FSF and PRIMo-Drain estimates of population and property value exposed to 1%- and 5%-annual-chance hazards diverge at finer scales of governance, for example by 4- to 18-fold at the municipal scale. FSF and PRIMo-Drain data often predict opposite patterns of exposure inequality across social groups (e.g., Black, White, Disadvantaged). Further, at the county scale, we compute a Model Agreement Index of only 24%—a ~1 in 4 chance of models agreeing upon which properties are at risk. Collectively, these differences point to limited capacity of FSF data to confidently assess which municipalities, social groups, and individual properties are at risk of flooding within urban areas. These results caution that national-scale model data at present may misinform urban flood risk strategies and lead to maladaptation, underscoring the importance of refined and validated urban models.

README: Comparing First Street Foundation and PRIMo Flood Hazard Data Across the Los Angeles Metropolitan Region

PI: Brett Sanders, University of California Irvine, Department of Civil Engineering (bsanders@uci.edu)
Data Questions Contact: Jochen Schubert, University of California Irvine, Department of Civil Engineering (j.schubert@uci.edu)
Other Authors:
Mach Katharine, University of Miami

Year Published: 2024
Years Data Collection: 2020-2024

Geography: Los Angeles, CA, USA

Description of the data file

Two data files are provided to calculate parcel level flood risk across Los Angeles:
parceldatatable.csv contains social data (e.g., population estimates, population fractions by race and ethnicity, and Neighborhood Disadvantage Index (NDI) values)
primo_flooddepth_table.csv contains flood hazard data generated by PRIMo and PRIMo-Drain

parceldatatable.csv contains the following variables:
id Dataset ID
apn Asessors Parcel Number
city Parcel City
tract Parcel Census Tract Number
blkgrp Parcel Census Block Group Number
blockid Parcel Census Block Number
popda Parcel Population
propvalue Parcel Value (US$)
ndi Neighborhood Disadvantage Index (UCI)
sovi Social Vulnerability Index (University of South Carolina)
svi Social Vulnerability Index (CDC)
fracasian Census Fraction non-Hispanic Asian
fracblack Census Fraction non-Hispanic Black
frachispanic Census Fraction Hispanic
fracwhite Census Fraction non-Hispanic White
medinc Block Group Median Income

Please note that parcels with zero population may exhibit "NaN" values (no Data) for population statistics. In some cases, zero population parcels may be attributed with population statistics. That is because statistics from the US Census and vulnerability indexes are downscaled from Tract or Block Group geographies, and zero population parcels may intersect those larger geographies. Vice-versa, it can also occur that parcels with a non-zero population are attributed with "NaN" population statistics due to their intersection with a larger Tract geography for which no vulnerability index was calculated.

primo_flooddepth_table.csv contains the following variables:
id Dataset ID
apn Asessors Parcel Number
h_100 PRIMo 1%-annual-chance flood depth
h_100_drain PRIMo-Drain 1%-annual-chance flood depth
h_20 PRIMo 5%-annual-chance flood depth
h_20_drain PRIMo-Drain 5%-annual-chance flood depth

For flood risk analysis use the included Matlab (.m) codes:
inequality_exposure.m calculates exposed populations
inequality_propvalue.m calculates exposed property
municipal_exposure_propvalue.m calculates Gini indices which quantify social inequalities

License:

For usage and publication with the above data please contact: Prof. Brett Sanders - bsanders@uci.edu

Funding

National Science Foundation, Award: HDBE-2031535

National Oceanic and Atmospheric Administration, Award: NA16NOS4780206