The paper deals with a very interesting and innovative topic, which is the use of LiDAR and SCADA data for a model-based estimate of mooring lines damage equivalent loads. The research design is accurate and the simulations are extensively and clearly described. The results are well presented and coherent.

I understand that, given the innovativeness of the work, there are not so many references against which comparing. Anyway, I think that a deeper discussion about how good the results of Figure 10 are could improve the manuscript. Similarly, I am interested in further insights about how the results and the importance of the various features might change when dealing with measured, instead than simulated, data.

Finally, the most important remark I have on the paper is that in my opinion it is fundamental to give the reader as soon as possible the information that the manuscript deals with simulated data. Therefore, I suggest to change the title in something like "Machine learning based virtual load sensors for mooring lines using simulated motion and lidar measurements".

The paper addresses core knowledge regarding influence of environmental conditions on mooring line loadings. The title is indeed misleading as suggested by another peer-review (e.g., Does it only estimate fairlead tension time series + DELs or also mooring line tensions). The relevant scientific questions are presented with specific details. The structure is well formed, with relevant figures and tables to describe the findings. 



I did miss some validation steps, since the load sensors are virtual, the machine learning models have high uncertainty in themselves and using a virtual lidar is only able to replicate what we understand about the atmosphere. I have provided the relevant comments at specific locations, where I am missing the link or need more information and I hope the author is able to revise the manuscript to accommodate these requests.