Well the shock model by WHT can model this the problem is the data. I think Web has nailed something very fundamental for modeling complex systems.

Briefly when I used to model chaotic systems we worked very hard to make sure initial conditions where random in all directions. In my opinion the beauty of what Web has done is recognize that you can shape the initial conditions such that they are not evenly distributed thus the input effects the output.

This shaped pulse is in a sense a simple model for a complex system !

However this leaves open the problem of how to shock the model. Dispersion is and obvious first order shock but then what ?

And next we simply don't have the data.

It might be possible to take the raw data and guess the shocks something along the lines of a Fourier transform.

I don't know if web has looked at deconvolution i.e run the shock model in reverse.

As and example instead of using discovery data to mean production use it to mean capacity. And although the data is incomplete it does work to constrain the viable solutions. And of course we have hard data from above ground events. How the system rebounds from and above ground event helps to narrow the possible outcomes.

Web of course has known this all along but it just recently hit me that the shock model is fundamentally valid for modeling complex systems.