Normalizing is mainly carried out for comparison between different data set. For example, if you want compare the steepness values between two rivers, a normalized steepness value is more apt than just steepness index.

Thanks Mr. Ali Yunus about your answer

But I ask about the procedure in the same river basin, and I ask about the advantages and applications in river morphometry.... otherwise can I use it in any geomorphic index.

I would like to help you, but could you provide a little more detail?

you need to have some points with certain interval that have elevation values. then, you can import them into a spreadsheet application. there you build a xy chart after you have converted them into percentages. in one axis you have elevation on the other horizontal distance

Joe Wheaton (now at Utah State I think) has done some fantastic work on this topic.  I think one short answer to your question is that detrending the data allows for integration with floodplain DEMs and thus spatial hydrologic or other habitat modeling.  I'm not trying for self-promotion here.  But, you might find the attached paper interesting and helpful.

cheers,

Ray