Impact of spatial configuration of training data on the performance of Amazonian tree species distribution models

Members of the Amazon Team have just published a new paper!

The paper Impact of spatial configuration of training data on the performance of Amazonian tree species distribution models was based on the last chapter of Pablo`s PhD thesis and co-written with Kalle Ruokolainen, Jasper Van doninck and Hanna Tuomisto.

The authors modelled the distribution of five tree taxa at local scales using remote sensing variables and assessed how species distribution models (SDMs) differ when different spatial configurations (compact to elongated) are trained. They found that more elongated configurations of the training data produced more robust SDMs. Their findings are key when tropical forest management decisions are made based on SDMs.

Read the full study here.

Congratz to the authors!