Improve the uncertainty computation methods and approach

Overview

The current implementation of uncertainty involves modifying the ISO method for midsection measurement and modifying the IVE approach based on what is done for midsection hydroacoustics. These are built as classes in the uncertainty.py file. It's a good start, but there is much room for improvement.

Missing

There are hard-coded uncertainty components in both the ISO and IVE methods. The ISO method is correctly applied, as all of the necessary aspects are covered as table lookups in the documentation of the method. IVE has holes and is very likely not being used correctly. In both methods, we can better estimate errors for the various uncertainty contributions. I've added thoughts for each contribution below.

Uncertainty Contributions and notes (will update)

• systematic (u_s)
• instrument repeatability (u_c)
• uncertainty in depth at 1-sigma (u_d)
  • @tknight : is any of this included in AC3?
• uncertainty in mean velocity (u_v)
  • we can "wiggle" the velocity in a few ways
    • change the search line distance
    • change the flow angle
    • compare the optimized flow angel to forced flow angel equal to the XS normal
    • move the grid node to the left/right of it's location and compare velocity
• uncertainty in width (u_b)
  • @tknight : is any of this included in AC3?
• uncertainty due to the number of verticals (u_m)
• uncertainty due to the number of points in the vertical (u_q)