autocalibration

parameter estimation and automatic calibration

dependencies

• optuna: for optimization
• modelskill: for evaluation of model skill
• mikeio: for altering roughness map and pfs

dependencies are included in the autocalibration environment.

structure

• data: folder where model data and observations to calibrate to reside.
  • Example_Corner: data for simple test case from MIKE FM HD examples "BedResistance". Added
• notebooks: folder with jupyter notebooks
  • example_corner: simple test case for FM HD where bed resistance map with two zones is modified

workflow for modification of roughness map

• run simulation with "base" roughness
• evaluate model vs. observation (modelskill)
  • objective: optimize water level at two points
  • metric: RMSE, CC
• adjust roughness map (mikeio)
  • read template roughness map
  • identify zones based on values in template
  • alter roughness in zones
• start from beginning with altered roughness map and repeat