Describe the bug
When creating the nodal attribute Mannings n landcover parameter, we often have to partition the domain into chunks/slices to fit into available RAM. When this happens, sometimes N>1 cell-averaging stencil overlaps partitions and values can become duplicated leading to errors in the fort.13 file creation for a model such as ADCIRC.

Onur could you please paste here the latest mannings n function we were working on so no one forgets for the next time?