{chmloader}: Let's compare notes!

[h-a-graham]

🌳🎋🌲🌴👋 Welcome! 👋🌳🎋🌲🌴

We’re using "Discussions" as a place to talk about the recently released Canopy Height Model (CHM)
dataset by WRI and Meta (Tolan, et al. (2024). We see huge potential in this dataset and the model used to create it and want to build a more quantitive sense of how the model performs across different environments and geographies.

We have developed this package to make it easier to download the dataset and carry out some (somewhat rudimentary) comparartive assessments with your own (or other) canopy height models.

It would be wonderful if you would like to join this discussion here and share your thoughts and the results from running chmloader::compare_models on another canopy height model.

There is no need to share specific location data about your specific model, the package will automatically extract some very basic information about the continent and biome where it is located. For now we believe this is sufficient information for us to make some basic judgements about the forest type etc.

Please try to be constructive, we see this as a learning opportunity both for us as researchers but also for the data developers, it is undoubtedly a massive achievement to produce a CHM at the global scale and at such a fine resolution. We believe that this approach may well signal a step change in the way we model forest structure with remote sensing. Let's learn more about it together!

Also if you would like to submit issues or pull requests for improvements or new features, that would be most welcome 😀

Here is an example of how to generate a comparison plot, where we use one of the reference datasets included in {chmloader}

library(chmloader)
# load your chm data as a SpatRaster with terra::rast()
# here we are loading on of the {chmloader} example datasets
new_forest <- reference_data("new_forest")
# Now run the `compare_models()` function on the `new_forest` object
compare_models(
  r1 = new_forest,
  aggregate = c(5, 10),
  title = "New Forest, England",
  drop_zeros = TRUE
)
#> ℹ meta/WRI CHM not provided, downloading now...
#> ✔ CHM downloaded successfully!
#> [[1]]

#> 
#> [[2]]

#> 
#> [[3]]

# if you want access to the meta/WRI CHM model then you should first download
# it with `download_chm()` and then run `compare_models()` including it as the
# `r2` argument like so:
# myCHM <- terra::rast("path/to/your/CHM.tif")
# metaCHM <- download_chm(myCHM)
# compare_models(r1 = myCHM, r2 = metaCHM)

^{Created on 2024-05-01 with reprex v2.0.2}