Dear Diana,

I will try to clarify what is needed for the patch cover raster. This raster must be a floating point number with a minimum value of 0 and a maximum value of 1. The values should be 0 or NoData where the ID raster does not show reefs. Where the ID raster does show reefs, the value should be greater than 0, up to a maximum of 1. The value indicates the proportion of the raster that is covered by reef. For example, the value 0.5 means 50% and 0.9 means 90%.

The question often arises: what is the purpose of this raster? Why does the tool require it at all? The purpose of this raster is to allow the user to account for some cells not being completely covered by suitable habitat. At first, it may seem unnecessary. After all, why not just increase the resolution of the analysis, e.g. to 100x100 meter cells, and just mark cells having reef present or absent? That would simplify things. The problem is that method used by the tool is usually too computationally expensive to be used at such high resolutions, and ocean currents data are usually not available at that high resolution either. It is usually necessary to have a resolution of 5x5 km, 10x10km, 25x25 km or coarser. When using such coarse resolutions, it is often the case that some cells contain only a little reef, while others contain a lot. The patch cover raster helps with this problem, by allowing different cells to contain different amounts of suitable habitat.

The reason the MGET documentation mentions the ArcGIS Block Statistics tool is that Block Statistics can be a simple way to take high resolution reef data (e.g. 1x1 km) and summarize it into a coarse resolution (e.g. 25x25 km) patch cover raster. But you don’t have to use the Block Statistics tool. The main thing is that you need to have a raster ranging from 0 to 1. You can produce it by some other means than Block Statistics.

I hope that helps!

Best,

Jason