The problem at hand includes 3 shapefiles. 2 Land Use classified satellite images and 1 regular polygon shapefile that covers the area of a National Park in my country. The software used in this analysis is QGIS 2.14.7. All shapefiles are using the SIRGAS 2000 CRS (EPSG: 4674)
My assignment was rather simple, I had to calculate the area for each class on the Land Use inside the park. The mold's area (the shapefile with the boundaries of the national park) is of 870.000 hectares.
I made sure both satellite images didn't overlap and merged them together. After that, I clipped the merged shapefile using the boundaries shapefiles as the mold. The output shapefile was created with the exact shape of the mold. Apparently everything was ok.
At last, I use the statistics for categories tool (located on the toolbox) to get a small table with the sums of each class. When I sum all the sums, the area goes to approximately 938.000 ha.
Any clues on why this might have happened?
I tried the topology checker plugin. I found 20 errors close to the limits of the satellite images, but they are small and wouldn't make up for almost 60.000 ha. Is it possible that just by the presence of errors, QGIS goes cuckoo?
I know I could dissolve the satellite shapes, but I'm not on a very powerful computer and the processing takes too long. Would it make a big difference anyway?
They're all on the same CRS, including the project.