How to write the results of lidR grid_metrics to a georeferenced raster?

Question

I'm using the lidR package to calculate LiDAR grid metrics (p.18 documentation) for forestry applications. For example, the following calculates the mean elevation within a 20m cell:

library(lidR)

LASfile <- "/Users/aaron/Desktop/data/1958-08-53.laz"
lidar <- readLAS(LASfile)

l <- grid_metrics(lidar, mean(Z), 20)

I can plot the results using the following approach:

grid_metrics(lidar, mean(Z), 20) %>% plot

What is the proper method to write the grid metrics to a georeferenced raster (e.g. geotiff)?

Answer 1

In lidR package there is a function as.raster which is faster than rasterFromXYZ

r <- as.raster(l)

If l is several columns wide (more than one metric), the output is a RasterStack instead of a RasterLayer

Answer 2

Since the output is a dataframe with X,Y coordinates and a value (V1) column, you can use rasterFromXYZ from the raster package. Then write to GeoTIFF format using writeRaster, also from the raster package.

library(lidR)
library(raster)

LASfile <- "/Users/aaron/Desktop/temp/lidR_gridmetrics/1958-08-53.las"
lidar <- readLAS(LASfile)

l <- grid_metrics(lidar, mean(Z), 20) 

r <- rasterFromXYZ(l, crs='+proj=utm +zone=15 +ellps=GRS80 +datum=NAD83 +units=m +no_defs ')
writeRaster(r, filename="/Users/aaron/Desktop/temp/lidR_gridmetrics/1958-08-53.tif", format="GTiff", overwrite=TRUE)