I am fairly new to mosaic datasets and I haven't been able to figure out if what I'm attempting to do is appropriate for a mosaic dataset. I have constructed a mosaic dataset in ArcGIS 10.5, which is made up of rasters for 3 continents (Africa, Asia, South America/Latin America) at 5 different time points (2000, 2005, 2010, 2015, 2020).
Each of these rasters holds information on the number of people within each pixel. I'm interested in summing up population for both an area of interest and several time points of interest. To find my area of interest I've created a Reference Mosaic Dataset and set the spatial extent to my area of interest. This works fine. Here is my Reference Mosaic Dataset set to the extent of Kenya:
Then, to get my years of interest, I need to select each raster within the Reference Mosaic Dataset that is consistent with the year in question (note that the first screen shot doesn't capture a column that I've created for each Year). What I'd like to do is search my Reference Mosaic Dataset for each year and sum up the population within the area of interest. So, for my Reference Mosaic Dataset, I'd like to select first the raster for the Year 2000 and sum up the Kenyan population, then select the raster for the Year 2005 and again sum up the Kenyan population while keeping a running total (ie, 2000 population plus 2005 population). Ideally, given the size of these rasters, I'd like to somehow be working with only temporary rasters for each selection.
Is there an efficient way to iterate through the rasters within the Reference Mosaic Dataset where I select the first time point, perform an operation, save that result, then work with the next time point, and so on?