v0.7.0

Added

• Adding a new functionality to create a BehaviourLandCoverChange object from shapefiles of glacier extents.
• Adding a function to compute potential clear-sky direct solar radiation as defined by Hock (1999).
• Adding a temperature index melt model based on Hock (1999).
• Adding the possibility to discretize the hydrological units based on the mean annual potential radiation used in Hock (1999).
• Adding an aspect-related degree day factor.
• Adding a reference prediction based on a bootstrapping approach of the observed discharge.
• Adding the option to define a parameter shared by different processes (e.g., snow and ice melt).
• Allow calibrating a single parameter set for multiple catchments.
• Adding a Results class to simplify outputs analysis.
• Adding maps plotting and spatio-temporal animations of the model outputs.
• Handling yearly recurrent forcing with a 'day_of_year' approach.
• Moving the model structure definition from C++ to Python.

Changed

• Transitioning to a dictionary-based structure definition.
• Moving the parameters generation out of the Socont class.
• Moving the process parameters and forcing definitions to respective classes.
• Changing workflow in load_from_csv.
• Restructuring examples.
• An error is raised if some parameter values are missing.
• The parameter estimation and evaluation functions are restructured to a new trainer.py file.
• Adding the date/time to the log file name.

Fixed

• Socont quick discharge has been fixed... For real, this time.
• Fixing the SpotpySetup initialization.
• Fixing issue with hydro unit initialization in catchment discretization.
• Fixing a bug that occurred when the study area outline comprised multiple polygons.
• Adapting the FixLandCoverFractionsTotal() function to also work with cases where the hydro unit is fully glaciated at the beginning of the Behaviour time series.