Recent fire history enhances semi-arid conifer forest drought resistance
Data files
Oct 28, 2024 version files 744.86 MB
-
control_south_sierra_FRAP_2km_buffer_300pt_5_fire_year_10tree_ts4_30m_20231204.csv
67.85 MB
-
control_south_sierra_Rx_2km_buffer_300pt_5_fire_year_10tree_ts4_30m_20231204.csv
63.77 MB
-
control_south_sierra_sev_2km_buffer_600pt_5_fire_year_20tree_ts4_30m_20231204.csv
74.05 MB
-
fire_south_sierra_FRAP_rxfire_300pt_5_fire_year_10tree_ts4_30m_20231204.csv
62.24 MB
-
fire_south_sierra_FRAP_wildfire_300pt_5_fire_year_10tree_ts4_30m_20231204.csv
67.79 MB
-
fire_south_sierra_USFS_sevfire_600pt_5_fire_year_20tree_ts4_30m_20231204.csv
71.80 MB
-
fire21_1_shp.zip
173.91 MB
-
Monthly_towerdata3_publish.csv
49.47 KB
-
README.md
9.74 KB
-
subsections_shp.zip
27.17 MB
-
UCIupwind_pixels_NDVI_met_30m_export.csv
5.05 MB
-
VebBurnSeverity18_1_shp.zip
131.17 MB
Abstract
Interaction effects from past climate change-amplified disturbances in fire-prone forests can produce ecological feedbacks that amplify or dampen subsequent disturbances. Climate change is increasing wildfire burned area, wildfire severity, and incidents of drought-induced forest dieback (widespread tree mortality). These climate change amplified disturbances threaten forest’s ability to regulate climate, provide water, and store carbon. Greater burned area makes interaction effects due to past fires more likely, which highlights the importance of understanding whether interaction effects produce negative or positive feedback on subsequent disturbances such as forest dieback. We constructed a forest chrono-sequence by combining a geospatial database of historical fire with a time series of Landsat satellite observations for forests in the Sierra Nevada of California to assess the impact of fire history on vegetation recovery, water use (evapotranspiration), and as feedback on subsequent forest dieback. We used these data sets to assess: 1.) How does a history of prescribed fire versus wildfire change forest cover and water use? 2.) How sensitive are forest structure and water use to fires of varying severity? 3.) How do these fire-induced changes to forest structure and water use impact forest dieback intensity? Forests with recent fire history had reduced tree cover, increased shrub cover, and decreased water use, with the greatest changes due to high-severity wildfires. Through 20 years post-fire water use neared pre-fire conditions across all fire types and severities, while decreased tree cover and increased shrub cover persisted following high severity wildfires and gradually returned to pre-fire conditions for moderate and low severity fire. Fire history decreased forest dieback intensity compared to unburned controls, which suggests a dampening (negative) feedback. The reductions in forest dieback intensity were equal to or greater than reductions in tree cover and greater than reductions in water use, which suggests that post-fire reductions in tree cover combined with reductions in water use enhanced forest drought resistance more than either reduction individually. In fire-prone conifer forests, interaction effects from past fires will dampen subsequent drought-induced forest dieback, introducing novel ecological feedback with implications for forest management and climate mitigation.
README: Recent fire history enhances semi-arid conifer forest drought resistance
These data sets and scripts allow for the creation of all figures and supplementary figures and tables cited in
Norlen, C.A.; Hemes, K.S.; Wang, J.A.; Randerson, J.T.; Battles, J.J.; Tubbesing, C.L.; Goulden, M.L. (2024) "Recent fire history enhances semi-arid conifer forest drought resistance". Forest Ecology and Management
Data Access
The data sets required to create the figures are available in the following DRYAD repository:
Norlen, C.A.; Hemes, K.S.; Wang, J.A.; Randerson, J.T.; Battles, J.J.; Tubbesing, C.L.; Goulden, M.L. (2024). Recent fire history enhances semi-arid conifer forest drought resistance. Dryad Data Repository. https://doi.org/10.5061/dryad.s4mw6m9f2
Description of the data and file structure
Shape File of USFS Ecological Subsections used to create Manuscript Figures 1a, S1
- subsection_shp.zip
Shape Files of CALFIRE fire perimeters used to create Manuscript Figures 1a,c, S1a,b
- fire21_1_shp.zip
Shape Files of USFS fire perimeters with added fire severity information used to create Manuscript Figures 1c, S1c
- VebBurnSeverity18_1_shp.zip
CSV files with samples from CALFIRE wildfire and prescribed fire perimeters (burned) and 2-km buffers (controls). Each time series sample for the data set is in a row of each CSV file The data includes Evapotranspiration (AET, mm/yr), annual precipitation (ppt, mm/yr), annual temperature (tmax, C), climate normal precipitation (clm_precip_sum, mm/yr), climate normal temperature (clm_temp_mean, C), ADS dieback (tpa, dead trees/acre), NDMI (unitless),
Wang et al, 2022 Vegetation Cover: Tree Cover (Tree_Cover, % cover), Shrub Cover (Shrub_Cover, % cover), Herbaceous Cover (Herb_Cover, % cover), Bare-ground Cover (Bare_Cover, % cover), Allred et al, 2021 Vegetation cover: Tree Cover (TRE, % cover), Shrub Cover (SHR, % cover), Annual Forbs and Grasses (AFG, % Cover), Perrennial Forbs and Grasses (PFG, % cover), Bare-ground (BGR, % cover), elevation (meters), latitude, longitude, forty-eight-month standardized precipitation index (SPI48, unit less), LANDFIRE Existing Vegetation Type (unitless), most recent fire year in 2010 (fire_year_2010), most recent fire year in 2019 (fire_year_2019), most recent fire year in 2020 (fire_year_2020), most recent fire type in 2010 (fire_type_2010), most recent fire type in 2019 (fire_type_2019), most recent fire type in 2020 (fire_type_2020), number of fires record as of 2010 (fire_count_2010), number of fires record as of 2019 (fire_count_2019), number of fires record as of 2020 (fire_count_2020), Column number for each sample (system.index, unitless), Stratification Layer (stratlayer, unitless) for each grid cell. There are multiple columns for each variable with pre-fixes from X1_ (1985) to X35_ (2019) to indicate years. The data is used in the following script: fig2_fig_4_figs1_table_1_table_s1_frap_rx_fire_recovery_ts_30m.r to create Figures 2, 4, 5, S2, S4, S9, S11, R3, and Tables 1, S1.
- fire_south_sierra_FRAP_wildfire_300pt_5_fire_year_10tree_ts4_30m_20231204.csv (wildfire burned samples)
- control_south_sierra_FRAP_2km_buffer_300pt_5_fire_year_10tree_ts4_30m_20231204.csv (wildfire control samples)
- fire_south_sierra_FRAP_rxfire_300pt_5_fire_year_10tree_ts4_30m_20231204.csv (prescribed fire burned samples)
- control_south_sierra_Rx_2km_buffer_300pt_5_fire_year_10tree_ts4_30m_20231204.csv (prescribed fire control samples)
CSV files with samples from USFS wildfire perimeters with fire severity information (burned) and 2-km buffers (controls). Each time series sample for the data set is in a row of each CSV file The data includes Evapotranspiration (AET, mm/yr), annual precipitation (ppt, mm/yr), annual temperature (tmax, C), climate normal precipitation (clm_precip_sum, mm/yr), climate normal temperature (clm_temp_mean, C), ADS dieback (tpa, dead trees/acre), NDMI (unitless), Wang et al, 2022 Vegetation Cover: Tree Cover (Tree_Cover, % cover), Shrub Cover (Shrub_Cover, % cover), Herbaceous Cover (Herb_Cover, % cover), Bare-ground Cover (Bare_Cover, % cover), Allred et al, 2021 Vegetation cover: Tree Cover (TRE, % cover), Shrub Cover (SHR, % cover), Annual Forbs and Grasses (AFG, % Cover), Perennial Forbs and Grasses (PFG, % cover), Bare-ground (BGR, % cover), elevation (meters), latitude, longitude, forty-eight-month standardized precipitation index (SPI48, unit less), LANDFIRE Existing Vegetation Type (unitless), number of fires record as of 2010 (fire_count_2010), number of fires record as of 2019 (fire_count_2019), number of fires record as of 2020 (fire_count_2020), most recent fire severity in 2010 (fire_sev_2010), most recent fire severity in 2019 (fire_sev_2019), most recent fire severity in 2020 (fire_sev_2020), most recent fire ID in 2010 (fire_ID_2010), most recent fire ID in 2019 (fire_ID_2019), most recent fire ID in 2020 (fire_ID_2020), Column number for each sample (system.index, unitless), Stratification Layer (stratlayer, unitless) for each grid cell. There are multiple columns for each variable with pre-fixes from X1_ (1985) to X35_ (2019) to indicate years. The data is used in the following script: fig3_fig5_fig_s2_table_2_table_s2_fire_sev_recovery_ts_30m.r to create Figures 3, 6, 7, S3, S5, S8, S10, S12, R4, and Tables 1, S2,
- fire_south_sierra_USFS_sevfire_600pt_5_fire_year_20tree_ts4_30m_20231204.csv (burned samples)
- control_south_sierra_sev_2km_buffer_600pt_5_fire_year_20tree_ts4_30m_20231204.csv (control samples)
CSV files of annual data from 10 eddy covariance tower locations were used to create Figure S4. Each file is for one of the 10 eddy covariance sites. Each file contains the following variables: wYEAR (water year), Evapotranspiration (ET, mm/yr), n_days (number of days with data), and ID (site description).
CSV file of system index (system:index), annual NDVI (NDVI_mean, unit less), Pixel # year, Pixel # (Number of 9 upwind Landsat pixels from tower), Site (FluxNext Site ID), precipitation (ppt, mm/yr), solar radiation (srad, W/m^2), temperature (tmean, C), year and month (yearmonth), geographic coordinates (.geo) extracted from PRISM, GRDMET, and Landsat for each of the 10 eddy covariance sites used to create Figure S6, S7.
- UCIupwind_pixels_NDVI_met_30m_export.csv
CSV file of FluxNext Site (Site), date (Mean date), daily ET (Efill.1, mm/day), Predicted ET (Predicted ET, mm/day), Dry Season Drawdown (Bucket 2 DSD, mm), Measured ET divided by Predicted ET (MeasureE/Predicted E, unitless) from combined records of 10 eddy covariance sites.
- Monthly_towerdata3_publish.csv
Sharing/Access information
Data was derived from these publicly available sources:
- Landsat data on Google Earth Engine (GEE): https://developers.google.com/earth-engine/datasets/catalog/landsat
- 48-month Standardized Precipitation Index (SPI48) data: https://wrcc.dri.edu/wwdt/archive.php
- California state perimeter: https://developers.google.com/earth-engine/datasets/catalog/TIGER_2016_States
- California Fire Perimeters: https://frap.fire.ca.gov/mapping/gis-data/
- LANDFIRE Existing Vegetation Type: https://www.landfire.gov/viewer/
- USFS Ecological Subsections: https://data.fs.usda.gov/geodata/edw/datasets.php?xmlKeyword=ecomap+subsection
- USFS Aerial Detection Surveys: https://www.fs.usda.gov/detail/r5/forest-grasslandhealth/?cid=fsbdev3_046696
- PRISM Precipitation data: https://developers.google.com/earth-engine/datasets/catalog/OREGONSTATE_PRISM_AN81m
- PRISM Temperature data: https://developers.google.com/earth-engine/datasets/catalog/OREGONSTATE_PRISM_AN81m
- USGS Digital Elevation Model: https://developers.google.com/earth-engine/datasets/catalog/USGS_NED.
- SRTM Digital Elevation Model: https://www.usgs.gov/centers/eros/science/usgs-eros-archive-digital-elevation-shuttle-radar-topography-mission-srtm-1
- GRIDMET Solar Radiation: https://developers.google.com/earth-engine/datasets/catalog/IDAHO_EPSCOR_GRIDMET
- Eddy Covariance Data: https://www.ess.uci.edu/~california/
Code/Software
The code shared with this submission was written in R 4.3.3 and run using RStudio.
The code requires the tidyverse, sf, RSQlite, rFIA, RSToolbox, patchwork, ggpubr, kableExtra, and gstat packages.
R script was used to create Figure 1, S2 of the manuscript.
- fig1_fire_history_maps.r
R script used to create Figures 2, 4, 5, and Table 1 of the manuscript and Figures S2, S4, S9, S11, R3, and Table S1.
- frap_rx_fire_recovery_dieback_comparison_30m.r
R script used to create Figure Figures 3, 6, 7, and Table 1 of the manuscript and Figures S3, S5, S8, S10, S12, R4, and Tables S2,
- fire_sev_recovery_dieback_comparison_30m
R script used to create Figures S6 and S7.
- fig_s9_ndvi_et_scaling