This talk addresses some issues of optimal control of systems whose behavior is described by partial differential equations. One of the motivations is comparing experiments and simulations, a central aspect of program validation.
Another motivation involves state and parameter estimation for physical systems, even in the presence of bad data; here the drivers have been atmospheric modeling and safe, efficient pipeline operation.

Various parts of the work described here is joint with Andrei Draganescu (University of Maryland Baltimore County), Henry H Rachford, Jr., and Richard Carter (GL Noble Denton).