I have a raster DEM and a reclassified raster DEM with different elevation ranges. I have a fishnet grid box (polygon). What I need to do is calculate the mean values of data lying inside those ranges for each grid. Say, I have set 3 elevation ranges 0-500, 500-1000 and 1000-8879m. Now for each grid, I need the mean of values of raster DEM lying inside those ranges. For eg. 2, 43, 24 lying in first elevation range should give their average value for a grid cell. Similarly, two other values are to be found accordingly for the other ranges. Hence I will have three values for a grid. It is possible that some grids may have only one or two values. The raster is 90m and grid is 25km resoultion.
I would think that you could use a Zonal Statistics Analysis for this. I would calculate Zonal Statistics first for the values within each range using the DEM as the Value data and the reclassified Raster as the zone data. Then you could do a second Zonal Statistics run using the fishnet grid as the zones and the previous zonal statistics as the data. Actually the second analysis could just be done using Tabulate Areas and that may give you a more relevant result. That would give you mean DEM values for each range of values in the reclassified DEM within the fishnet grid.