Climate matching by drawing a buffer around occurrence points

[soriadelva]

At the moment, we keep points with every coordinate uncertainty level and also those that are close to the border with a different climate zone. To minimize bias related to this, draw a buffer around each occurrence point with the radius being equal to the coordinate uncertainty (i.e., records with a higher uncertainty will get a higher buffer around them). Then, calculate the amount of overlap between this buffer and the climate zone layer of the time period in which the observation was made. Based on this % overlap, calculate the presence of the point in each climate zone as follows: 60% overlap of occurrence record buffer with zone A and 40% with zone B = 0.6 occurrence records in zone A and 0.4 in zone B.

In other words: n_climate (zone A) = Area of buffer in climate zone A / Total area of buffer (and this for all occurrence records of the species of interest)

Special attention for the following situations:

• If a buffer overlaps with the marine environment, keep only the terrestrial overlap, recalculate the buffer surface ...
• If a buffer overlaps with the buffer of another point: chances will be lower due to thinning but this may still occur when coordinateUncertainties are large, in that case don't merge buffers, calculate it separately for each point as if the other weren't there.

[SanderDevisscher]

I would include some logic to, if a buffer overlaps also with the marine environment, keep only the terrestrial overlap, recalculate the area and the occurrence record buffer.

[soriadelva]

I would include some logic to, if a buffer overlaps also with the marine environment, keep only the terrestrial overlap, recalculate the area and the occurrence record buffer.

Yes, I agree!