Eaton Fire Lidar Analysis: Pre- and Post-Fire Comparisons

The Eaton Fire left a significant mark on the landscape, burning nearly 10,000 structures and surrounding vegetation. Lidar News reported a few weeks ago about the availability of lidar data for the Palisades and Eaton fires. Fortunately, lidar data was collected in the area of the Palisades and Eaton fires both before (2018) and promptly after the devastating fires. This provides many opportunities to compare data from before and after those fires. Here, we chose to use lidR, a package within the R Project for Statistical Computing to provide a simple visualization of a digital surface model before and after the fire.

Data was accessed from the USGS The National Map website. To clarify data availability, pre-fire data from 2018 included a point cloud and a digital surface model for the extent of the Eaton fire. Post-fire, the point cloud itself was not available.

The basic methodology was to access the LAZ files, mosaic them, generate a DSM, create a hillshade, and compare pre- and post-fire data. Generating a DSM from the LAZ files for the pre-fire data was not necessary, as a DSM is available from USGS The National Map, but we include this process for reference.

Pre-fire Digital Surface Model from within the extent of the Eaton fire. Numerous houses can be observed.

Post-fire Digital Surface Model of the same extent. The vast majority of structures appear to be lost.

We also uploaded these images into the King Lab Juxtapose software to allow for a more dynamic comparison


The R code is provided below for your reference.

# Call needed packages & Import files -------------------------------------

# packages
library(lidR)
library(raster)
library(terra)

# 2025 DSM file path
file_path_2025 <- "/Users/nathanroe/Downloads/post_fire_dsm_11SLT003945037845.tif"

# Create terra SpatRaster for 2025 DSM
post <- terra::rast(file_path_2025)

# 2018 pre-fire LAZ file paths
file_paths_laz <- c("/Users/nathanroe/Downloads/USGS_LPC_CA_LosAngeles_2016_L4_6519_1894a_LAS_2018.laz",
                "/Users/nathanroe/Downloads/USGS_LPC_CA_LosAngeles_2016_L4_6519_1889b_LAS_2018.laz",
                "/Users/nathanroe/Downloads/USGS_LPC_CA_LosAngeles_2016_L4_6514_1889c_LAS_2018.laz",
                "/Users/nathanroe/Downloads/USGS_LPC_CA_LosAngeles_2016_L4_6514_1894d_LAS_2018.laz")

# Read LAZ files into list
las_files <- lapply(file_paths_laz, readLAS)

# Generate DSM from pre-fire LAZ ------------------------------------------

# Merge the LAS files using rbind
las_combined <- las_files[[1]]
for(i in 2:length(las_files)) {
  las_combined <- rbind(las_combined, las_files[[i]])
}

# Generate DSM using non-ground points
pre <- rasterize_canopy(las_combined, res = 1, algorithm = dsmtin())

# Create hillshades -------------------------------------------------------

# Pre-fire hillshade generation
pre_slope <- terrain(pre, "slope", unit="radians")
pre_aspect <- terrain(pre, "aspect", unit="radians")
pre_shade <- shade(pre_slope, pre_aspect, 40, 270)

# Post-fire hillshade generation
post_slope <- terrain(post, "slope", unit="radians")
post_aspect <- terrain(post, "aspect", unit="radians")
post_shade <- shade(post_slope, post_aspect, 40, 270)

# Ensure the same coordinate reference system is used
pre_shade <- project(pre_shade, terra::crs(post_shade))

# Crop pre-fire hillshade to post-fire hillshade
pre_shade <- crop(pre_shade, ext(post_shade))

# Plot! -------------------------------------------------------------------

plot(pre_shade, main="Eaton Pre-Fire", col=gray.colors(100))
plot(post_shade,main="Eaton Post-Fire", col=gray.colors(100))

Thanks for reading to the end!