Assess the Effect of Time Sampling

[WardLT]

Our measurements are in storage per data point. We should estimate the frequency needed to store data and still compute needed quantities (e.g., capacity) with required accuracies.

[WardLT]

Equally-spaced points may not be the right way of considering storage. Maybe we could define a limit of the curvature in the property to put bounds on the errors of linear interpolation for the data

[WardLT]

Suggestion from Paul Gasper: Keep the points at which we change steps