Various vector polygons for annual crop type map?

Hi again,

I’m planning to produce an annual crop type to monitor the advance of oil palm and annual crops on the upland forests. (from 2015-present)

Is it ok to use a single vector polygon/in situ shape file for each of the annual crop type map? Or do I need to collect features and made various vector polygons for each of the year?


I’m not sure I understand the question. If you prepare a different set of training data for each year, how will they differ from the previous ones?

If some of the parcels have changed their usage, the shapefile should be updated. Otherwise – if the data is still valid – I don’t think there’s any reason to change it.

What I mean is, is it ok to have a single set of vector polygons in producing a series annual crop type map? Or do I need to have a set vector polygon for each of my annual crop type map? How do these options affect my annual crop type results?

I’m guess that there were significant land use/cover changes om area of interest based on my high resolution images of Google Earth.

Then it sounds like keeping the same set of polygons (without changing their attributes) would mean providing wrong data to the processor for the latter years; that doesn’t seem like a good idea.

It might be better to keep the polygons fixed where possible, to make it easier to compare the map for one year with the map for the next one. But it’s more important that the training data is correct.