Adjust weights for PESTPP based on observation type

[jzwart]

Leaving the weights of the different observation types (e.g. temperature obs, flow obs) causes bias towards high value observations (or residuals) in objective function. According to PESTPP4.2.4 manual

A simple procedure that often works well is to assign observations of different types to different observation groups, and to then use a utility program such as PWTADJ1 from the PEST suite to ensure that the contribution made to the objective function by each observation group is about the same as that made by any other observation group at the commencement of the inversion process.
When calibrating a surface water model, or when including contaminant concentrations as a component of the calibration dataset of a groundwater model, it may be useful for the magnitudes of weights to reflect the magnitudes of measurements.

[jzwart]

Leaving flow and temperature weights all at 1 reduced RMSE for flow (6.1 RMSE compared to 6.8 for uncal), but inflated RMSE for temperature (~40 RMSE).

Currently trying out adjusting weight by log normalizing between 0 and 1. This gives more weight to low values (temperature values and low flow). See log normalizing (red) vs. linear (black) below

zoomed in x-axis

[jzwart]

Log normalization still biases flow. Setting temp weight to 1 and flow weights to 0.05 still biases calibration towards flow. Probably best to calibrate flow and temp separately for now.