Is it okay to do analysis on coarse rainfall data, since the area calculations can be highly misleading

I need to find the area of different classes within the rainfall data, which will be pixel-based, and since coarser data may under and over cover the study area (image below), the area calculations can be wrong. How to deal with it? My professor suggested this: "You can create a buffer of one pixel around India boundary so that you have data until India boundary if you clip the images with the buffered boundary."

but still, this won’t solve the problem as still, the coverage problem will exist.

How to deal with it?