Leopold equation I=Qp/CA  where I = intensity, Qp = peak Q, C =                 
runoff coefficient, A = drainage basin area (hugh used a C =                    
0.9 since dealing with very high hypothetical intensities, near                 
surface saturation will likely be rapid).                                       
                                                                                
As we discussed previously in the field, I have not only clast size             
issues in this regard, but also clast shape issues (e.g. flaggy in              
Fern, Equant @ LRIV).  Costas' (1983) theoretical method 1 "Helley              
(1969) deals with clast shape issues using short, int. and long                 
axis diameters in the velocity estimates... seems like this would               
be preferable to Costa's final regression equation.                             
                                                                                
Have you played with Costa "theoretical method 1"?                              
                                                                                
do you have other ideas that I should know about?                               
                                                                                
Do you see major problems with this analysis (before I could through            
the effort, which will be moderately time consuming)?                           
                                                                                
If you have played with method 1, I'm having trouble reproducing                
Costa's data and wondering if there are typos / errors in the published         
equations... have you run into this previously?                                 
                                                                                
                                                                                
I will also contact hugh about this                                             
