growR 1.3.0

Added

An additional option for determination of growing season start is available through ModvegeSite$set_SGS_method. This comes with a different available definition of the temperature sum and is reflected in the new methodsstart_of_growing_season,start_of_growing_season_mtd,weighted_temperature_sum`.

j_start_of_growing_season is written to output file headers.
Output file header contains cut_DOYs.
growR_run_loop allows use of previous year’s state variables as initial conditions for current year by means of the independent argument.
stubble_height is now a parameter (i.e. a field of ModvegeParameters) instead of a field of MovegeSite.
Lower biomass limit is only applied after ST2.

Window sizes in multicriterial thermal definition were incorrect.
Calculation of fAgeDR was incorrect.
Biomass is correctly updated after cuts.
dBM can become negative, as it should.
fW now uses correct threshold value of 3.8 instead of 3.81.
Correction of AEv calculation.
NULL value for simulations without cut_DOYs still produces a header in the output file.
Bug where error message was not printed correctly in case of missing parameters.
Biomass (BM, BMxx) was unaligned with biomass change (dBM, dBMxx).

Green biomass has a lower bound defined by a minimum grass height. This is to prevent the population from completely dying out and effectively simulates reserves.

The start of the growing season is now determined by the multicriterial thermal definition as described in chapter 2.3.1.3 of the dissertation of Andreas Schaumberger: Räumliche Modelle zur Vegetations- und Ertragsdynamik im Wirtschaftsgrünland, 2011, ISBN-13: 978-3-902559-67-8

Input integrity checking for WeatherData and ManagementData.
PScanPlotter (plot_parameter_scan) now allows to select which variable to display (from dBM, cBM and cBM_tot).
Instances where simulation results for variable dBM are visualized or compared to experimental data now allow specifying and integration window for dBM, which should be set to match the experimental reality. This concerns ModvegeSite$plot() and analyze_parameter_scan() as well as the compare.R script.
S3 dispatch for plot method of ModvegeSite objects -> plot(mvs) is now possible if mvs is a ModvegeSite instance.
ParameterData input checking: throws error on duplicate input parameter name.
Debug utility conveniences browse and browse_end.
ModvegeSite$plot_XXX functions for more insights into model behaviour, with XXX in water, limitations, growth and plot_var for generic variable plotting.

Input data CSV files are now actual CSV files, instead of “semicolon-separated value” files.
ParameterData now assumes default values for most parameters. check_parameters now only throws an error if any of the really
required parameters are missing.
Number of days per year is now inferred from weather data instead of hardcoded to 365, allowing simulations to be run for years with incomplete data.
Crop coefficient (parameters$crop_coefficient) and senescence capping are now model parameters instead of being hardcoded.

autocut: get_annual_gross_yield was incorrectly hardcoded to return 1.
ParameterData$set_parameters now updates initial condition values, if applicable.