I am currently working on describing land cover for habitats where my study species is using the Multi-Resolution Land Characteristics (MRLC) Consortium National Land Cover Database. I am taking GPS points where these animals have been found and including 100m and 1000m buffers. The issue that I am having is that output for different points/buffers end up being different sizes. This is because the NLCD breaks down land cover into 30m pixels, so if the buffer includes even a small portion of one of these pixels that entire pixel will be included in the area. I am standardizing the areas to give percent cover, so I think this minimizes the issue. However, I do not want to disproportionately include some pixels that need not be included. Is there a way to resolve this issue and made it so the output is the same area for each GPS point?