your zenodo badge here

li_et-al_natcom_2023

The Influence of Climate Change on Flooding and Social Inequalities from Remnants of Hurricane Ida

Xue Li^1*, Michael F. Wehner², David R. Judi¹ and Robert D. Hetland³

¹ Earth Systems Science, Pacific Northwest National Laboratory, Richland, Washington, USA
² Applied Mathematics and Computational Research Division, Lawrence Berkeley National Laboratory, Berkeley, California, USA
³ Atmospheric Science & Global Change, Pacific Northwest National Laboratory, Richland, Washington, USA

* corresponding author: xue.li@pnnl.gov

Abstract

Previous research has demonstrated that tropical cyclone-related precipitation and flooding have been increased by anthropogenic global warming. Our work aims to quantify the contribution of climate change to the deadly flooding from the remnants of Hurricane Ida (2021) and its impact on human society. We developed an analysis framework that combines lessons from climate change attribution science, two-dimensional hydrodynamic modeling, and flood impact evaluation, including a social inequality index, to estimate remnants of Ida flooding and consequences responding to current (locally 1\degree C warmer) and future (another 1\degree C warmer) climate change. We find that an additional quarter to a half million people were exposed to flooding due to current and future climate change. The human influence on flood impacts was larger with deeper flood water depth ($\ge$ 1 m) than with shallow depths. Socially vulnerable populations were found to be disproportionately more affected, and climate change exacerbates this inequality.

Journal reference

Li, X., M.F. Wehner, D.R. Judi, & R.D. Hetland. The Influence of Climate Change on Flooding and Social Inequalities from Remnants of Hurricane Ida. Nature Communications. submitted.

Data reference

Both input and output data are available in the following reference:

Li, X. et al. (2023). Data from "The Influence of Climate Change on Flooding and Social Inequalities from Remnants of Hurricane Ida" [Data set]. MSD-LIVE Data Repository. https://doi.org/10.57931/2000435

Data are organized in subfolders by topic.

Input data

Dataset	Path
Hourly Stage IV QPE	prcp/original/*.tif
NLCD 2019	impact/landcover/nlcd/nlcd_2019_land_cover_l48_20210604.zip
NLCD 2019 Developed Imperviousness Descriptor	impact/landcover/nlcd/nlcd_2019_impervious_descriptor_l48_20210604.zip
LandScan USA 2020 - CONUS Night	impact/population/landscan/conus_night.tif
CDC/ATSDR Social Vulnerability Index 2020	impact/population/svi/SVI_US2020.zip
HUC4 domains exported from NHDPlus HR WBD	support/domain.shp
HUC12 polygons exported from NHDPlus HR WBD	support/huc12.shp
Water masks derived from NHDPlus HR	flood/water_mask/*.tif
High-Water Mark	flood/validation/hwm_312.csv

Additional political boundary datasets were aquired from Census TIGER/Line Shapefiles and HIFLD to draw basemaps in figures and can be found in the support folder.

Output data

Include intermediate and final output data:

Dataset	Path
Total precipitation in present scenario	prcp/present_prcp_total.tif
Peak flood depth for each domain and scenario	flood/RIFT_results/*h.tif
Peak rate of rise for each domain and scenario	flood/RIFT_results/*dhdt.tif
Peak total head for each domain and scenario	flood/RIFT_results/*he.tif
Flood characteristic summary (HUC4 and HUC12)	flood/huc*/flood.csv
Climate change contribution to flood characteristics (HUC4 and HUC12)	flood/huc*/diff.csv
Validation results by domain	flood/validation/hwm_rift.csv
Flood exposure by domain, scenario, depth threshold, and landcover type	impact/nlcd/summary.csv
Downscale population by domain	impact/population/downscaled/*.tif
Population exposure by domain, scenario, and depth threshold	impact/population/summary.csv
Gridded overall SVI by domain	impact/population/svi/*.tif
SVI summary for background population by domain	impact/population/svi_bg.csv
SVI summary for exposed population by domain, scenario, and depth threshold	impact/population/svi.csv

Reproduce my experiement

Follow the steps below to reproduce my work:

1. Download data from Data Reference and copy to the data subfolder. All workflow scripts expect data in reletive locations.
2. The RIFT model is in the process of becoming open source. So the RIFT output files are provided as inputs in subsequent data processing and analysis.
3. Run the following scripts in the workflow directory to re-create this experiment:

Script Name	Description
prcp.py	Aggregate hourly precipitation data to total precipitation for plotting
flood_huc4.py	Sumarise flood characteristics and cliamte-related changes at HUC4 level
flood_huc12.py	Sumarise flood characteristics and cliamte-related changes at HUC12 level
popDownscale.py	Downscale LandScan population grids from 3 to 1 arc-sec
exposure.py	Summarise land and population exposure
sviRasterize.py	Rasterize SVI to match downscaled population grids

Reproduce my figures and tables

Use the scripts found in the figures directory to reproduce the figures and tables used in this publication.

Script Name	Description
plot_table.R	Script to generate my figures and tables Preferably run in IDE (e.g., RStudio)

Major figures

Precipitation and flood simulation results for the present scenario. Total precipitation (a) and simulated flood characteristics averaged at HUC12 level, including (b) peak flood water depth, (c) peak rate-of-rise, (d) peak total head, and (e) percent of land area flooded.

Climate change contributions to flood characteristics for the past 14% scenario at HU12-level. The left panel shows the spatial pattern of differences due to reduced precipitation, and the right panel shows the correlation of differences to the intensity of present-day value.

Flood impacts and climate change contribution at different impact levels. Flood impacts were presented from (a) area of developed land and (b) population exposed to floodwater exceeding specific depth thresholds. One ft is approximately 0.3 m.

Illustration of concentration curves and social inequality indices. Panel a illustrates the concept of the designed indices, and panel b shows data derived from the present scenario analysis for people exposed to different levels of floodwater.

Social inequality of flood exposure and the role of climate change. Both the absolute (a) and relative (b) social inequality indices are included. Data were summarised for people exposed to deeper than 0.5 ft (≈ 0.15m, light blue) and 3 ft (≈ 1m, dark blue) of floodwater.