Matlab Spline Solved

OK, today I spend a few hours on solving the spline problem in Matlab. I re-wrote some codes so that the curves built by spline function looked good. This time the spline function was only applied to interpolate and extrapolate one curve: the Total Annual Mean Rate of Exceedance. Other curves of different Mws were constructed by the assumption of linear relationship between each two given points. Well, I personally don't like the approach used here, but I think it is a better method so far. My questions about the this curve-construction method include:

• Linear relationship between each two points. Is this assumption reasonable?
• Assumption of parallel lines either at points with smaller or greater PGAs than first or last PGA from USGS data. Why use parallel lines? Can't we extend the lines by spline function?
• Broken lines happened when the contribution from some Mw were zero at given PGA. How can we explain the broken line condition? Can we connect the line by spline function? Probably not.

Mmm.... I need more checking works to verify the results. Good luck to me!

by Niner 04052006