User Tools

Site Tools


start:hype_file_reference:state_save

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
start:hype_file_reference:state_save [2018/11/07 16:07]
rcapell
start:hype_file_reference:state_save [2021/03/11 15:31] (current)
cpers
Line 1: Line 1:
-====== state_saveyyyymmdd[HHMM].txt ======+====== state_saveyyyymmdd[HHMM] ======
  
 State variables can be saved to a file and later used for starting a model simulation from that same point. This can be useful to shorten the simulation time, e.g to skip repeatedly simulating a warm-up period, or to simulate several forecasts after running the model up to date.  State variables can be saved to a file and later used for starting a model simulation from that same point. This can be useful to shorten the simulation time, e.g to skip repeatedly simulating a warm-up period, or to simulate several forecasts after running the model up to date. 
  
-State-files are saved for the dates given by ''​outstatedate''​ in [[start:​hype_file_reference:​info.txt|info.txt]]. The files are written to the [[start:​hype_file_reference:​info.txt&#​model_options|resultdir]] folder. To use a state-file as a starting state ''​instate''​ is set in [[start:​hype_file_reference:​info.txt|info.txt]]. A state-file with the date(time) given by ''​bdate''​ is expected and used as starting state. The starting state file is expected to be found in the [[start:​hype_file_reference:​info.txt&#​model_options|modeldir]] folder. There is one file per time step with saved states: state_saveyyyymmdd[HHMM].txt yyyymmdd[HHMM] is the date(time) of the start of simulation (''​bdate''​). For daily time step only the date is used in the file name.+State-files are saved for the dates given by ''​outstatedate''​ in [[start:​hype_file_reference:​info.txt|info.txt]]. The files are written to the [[start:​hype_file_reference:​info.txt&#​model_options|resultdir]] folder. To use a state-file as a starting state ''​instate''​ is set in [[start:​hype_file_reference:​info.txt|info.txt]]. A state-file with the date(time) given by ''​bdate''​ is expected and used as starting state. The starting state file is expected to be found in the [[start:​hype_file_reference:​info.txt&#​model_options|forcingdir]] folder. There is one file per time step with saved states: state_saveyyyymmdd[HHMM].txt/tgz. yyyymmdd[HHMM] is the date(time) of the start of simulation (''​bdate''​). For daily time step only the date is used in the file name. State files can be in text format (state_saveyyyymmdd[HHMM].txt) or compressed with gzip by the tar archive software (state_saveyyyymmdd[HHMM].tgz). The compressed option is set in [[start:​hype_file_reference:​info.txt|info.txt]],​ and can be used for Windows10 and Linux.
  
-The first row of the state_save-file hold integer codes for what settings were used when creating the file. The settings are checked against the simulation that is started. Most of the settings must be the same for the simulation to start. For instance number of subbasins and classes are checked, as is some model options and time step length. Number of substances simulated (and their internal order) is checked, but it is possible to use a starting state created from a simulation with substances (e.g. NP) to start a simulation of ony discharge. It is not possible run a model of only discharge starting from a state created with "​substance"​ T2 though. That is an exception because a T2 simulation turn on lake and river ice calculations and related states. Updating with the AR-method is possible to turn on or off between simulations using starting states.+====Format of state_saveyyyymmdd[HHMM].txt file==== 
 +The first row of the state_save-file hold integer codes for what settings were used when creating the file. The settings are checked against the simulation that is started. Most of the settings must be the same for the simulation to start. For instance number of subbasins and classes are checked, as is some model options and time step length. Number of substances simulated (and their internal order) is checked, but it is possible to use a starting state created from a simulation with substances (e.g. NP) to start a simulation of ony discharge. It is not possible run a model of only discharge starting from a state created with "​substance"​ T2 though. That is an exception because a T2 simulation turn on lake and river ice calculations and related states. Updating with the AR-method is possible to turn on or off between simulations ​where some of them are using starting states.
  
  
start/hype_file_reference/state_save.1541603273.txt.gz · Last modified: 2018/11/07 16:07 by rcapell