I was able to exactly recalculate shear stress in some test faces. The hydraulic radius R is calculated as quotient of wetted area and wetted perimeter of the face. Then, friction slope is calculated between cell centers of the two adjacent cells of the face.
The results were completely same as those in hdf file. But still confused how RAS Mapper calculate and generate shear stress map. For example:
Face shear stress in 135278 is 47.5 N/m2
Face shear stress in 135279 is 17.9 N/m2
Face shear stress in 135278 is 21.5 N/m2
but !!!!
Shear stress in the face point 182411 is 146.1 N/m2 (read in the RAS Mapper). See also unusual overestimated area of the shear stress in the image.
