The manuscript discusses a novel approach to assessing the key factors that drive global fire activity. It is a data-driven model approach in which empirical relationships for a variety of predictor variables are applied. The skill of the set of SOFIA models is compared to a machine learning technique and a process oriented fire model in reproducing observed (using the GFED and CCI datasets) fire activity (burnt area serves as the chosen metric).
Being the topical editor I have followed the paper from the beginning. I have studied the first submission and the revised version of the manuscript and I have carefully studied the reviewers concerns. Unfortunately, the reviewers' opinion is divided. Taking into account the fact that it has taken an unusual amount of time to obtain the reviews I have decided to act as third reviewer. And as such I am satisfied that, on the one hand, the manuscript fits into the scope of GMD and fulfils the criteria required for publication (code availability etc.). The manuscript is well structured and well written, and the quality of the presentation has, in my opinion, increased in the revised version.
On the other hand, I think that this manuscript represents a timely contribution to an important contemporary field of research. It may not contribute a quantum leap in the science of global fire activity in natural environments but it does contribute a number of very useful insights and tools and outlines a new, useful approach that will help to promote knowledge in this important research field.
I have next to no comments with regards to changes to the manuscript (only one minor suggestion) and recommend publication in GMD.
SPECIFIC COMMENTS:
Figure R3: I suggest to reduce the horizontal scale (population density) or add a zoomed-in region to show the gradient in more detail. As it stands, the curve looks very much like a step function. |