The NOAH data table we are working with have just one row for water covered land, while the land use index array from the dataset differentiates between water and lakes. This extra land use index not contemplated in the data table leads to wrong plots.
Added this line in cross_tables function:
new_map[new_map == 21] = 17#21 stands for Lakes in land use index. It is assigned to 17 which is Water in MODIS NOAH data table.
The NOAH data table we are working with have just one row for water covered land, while the land use index array from the dataset differentiates between water and lakes. This extra land use index not contemplated in the data table leads to wrong plots.