Lucet Valentin

Overview

This is a very interesting and relevant project. I think it's great that you're tackling a real issue, and will hopefully be able to support the NGOs you're working with to optimize their focus. While you are working to improve habitat connectivity, you are also helping the NGOs themselves work more effectively together, so you are supporting connectivity both ecologically and within the organizations!

The project is well thought out and developed. Some simplifications have been made that may need further justification, but this is likely just a question of refinement as the project itself is very well formed.

Project Objectives and Questions

The objectives and questions are well defined. I might suggest a minor editing change if you plan to present this to your clients later - use a word other than nefarious - as this translates directly into "wicked, evil, criminal" etc., and is likely too extreme for what you're trying to convey. You probably want something like "adverse" or "ecologically damaging".

Data

It appears that you have all the data you'll need to complete the project. I wonder if there will be any issues with synchronizing the data. Will you have to modify any of it to make sure it is in the same resolution, or of the same type, so it is comparable? I'm sure you've thought of this. ** I read further and see that you spoke about standardizing the data into poly-lines, so the above comment is irrelevant! **

Regarding the land use data, is 2010 the most recent data you can access? Quite possibly it is, but there may have been significant land use changes in the last 8 years that you will not be able to analyse with the 2010 data. Would it be worth it to compare the 2010 land use data with data from Earth Engine, or some other data source, to check if there are any major land use changes that your 2010 data doesn't account for? Possibly you could do some kind of quick comparison, to make sure there are no significant changes you might be missing (maybe they built a mega-mall in the middle of an eco-corridor in 2014 that the 2010 data doesn't reflect!).

Methodology

The methodology is well designed and has clearly been very well deliberated, even up to the point of which values will be assigned to which pixels in the change raster.

The standardization step does raise some questions. You were asked about this in class, so it has already been at least partially addressed. You mentioned that your goal when standardizing the corridors is to maintain the "purest" data, that at the center of the corridors, and you acknowledged that you may lose some data when this method is applied to the "thicker" corridors that some of the NGOs provided. Is there any other way of maintaining the data included in these thicker "patches"? Conversely, is there any utility in maintaining this data? It might be that you just need to provide further justification for why losing this data will not significantly effect your analysis.

Possible problems

The standardization step is the only one I can see leading to possible analysis errors. It will likely be challenging dealing with data from these different sources, but you seem to have an excellent grasp on how to move forward. I'm looking forward to seeing the results.

Created for Advanced GIS for Natural Resource Management, in the McGill University Department of Natural Resource Sciences, Professor Jeffrey Cardille