In recent CONUS baseline testing, in particular for running Topmodel and Lumped Arid/Semi-arid Model (LASAM) model, in discussions with hydrologists, it has become apparent that there is a bit of intricacies for running these two models. The intricacies involve SoilMoistureProfile (SMP)submodule. For this submodule, different initial configuration files are needed to run different hydrological models. For running CFE, we have used the config_conceptual.txt equivalent of the initial config, which is appropriate. We have generated the initial config appropriately for each catchment at CONUS scale. For running Topmodel, config_topmodel.txtversion of initial configuration should be used. In our testing, due to lack of initial configurations in this form on CONUS, we used the same initial configs as for CFE for SMP submodule with any needed input parameters for Topmodel initialized in the Sloth submodule. The computations run fine but the results probably can only be considered approximate. Using a single config_topmodel.txt for the SMP to run Topmodel appears to run only a few time steps, then abruply stops. For running LASAM, the appropriate version of initial configuration for SMP is config_layered.txt. Due to unavailability of accurate initial config for each catchment at CONUS scale, we used the same initial config for all catchments on CONUS. Nevertheless, This is an initial test of feasibility and in the process we debugged some module codes (such as memory leaking of LGAR-C) and uncovered the limit of some parameters beyond which some submodules would not work (ref. NextGen_on_CONUS in ngen documentation for more details). Despite the issues remaining as described above, this work has forged a path in performimg computations on such large scale, which makes future such computations within Ngen Framework more straightforward.
Current behavior
Unrealistic SMP initial configs. for Topmodel
Unrealistic SMP initial configs. for LASAM model
Expected behavior
Individually fitted SMP initial config. for each catchment at CONUS scale for Topmodel and LASAM.
In recent CONUS baseline testing, in particular for running
TopmodelandLumped Arid/Semi-arid Model (LASAM)model, in discussions with hydrologists, it has become apparent that there is a bit of intricacies for running these two models. The intricacies involveSoilMoistureProfile (SMP)submodule. For this submodule, different initial configuration files are needed to run different hydrological models. For running CFE, we have used theconfig_conceptual.txtequivalent of the initial config, which is appropriate. We have generated the initial config appropriately for each catchment at CONUS scale. For running Topmodel,config_topmodel.txtversion of initial configuration should be used. In our testing, due to lack of initial configurations in this form on CONUS, we used the same initial configs as for CFE for SMP submodule with any needed input parameters for Topmodel initialized in the Sloth submodule. The computations run fine but the results probably can only be considered approximate. Using a singleconfig_topmodel.txtfor the SMP to run Topmodel appears to run only a few time steps, then abruply stops. For running LASAM, the appropriate version of initial configuration for SMP isconfig_layered.txt. Due to unavailability of accurate initial config for each catchment at CONUS scale, we used the same initial config for all catchments on CONUS. Nevertheless, This is an initial test of feasibility and in the process we debugged some module codes (such as memory leaking of LGAR-C) and uncovered the limit of some parameters beyond which some submodules would not work (ref. NextGen_on_CONUS in ngen documentation for more details). Despite the issues remaining as described above, this work has forged a path in performimg computations on such large scale, which makes future such computations within Ngen Framework more straightforward.Current behavior
Unrealistic SMP initial configs. for Topmodel Unrealistic SMP initial configs. for LASAM model
Expected behavior
Individually fitted SMP initial config. for each catchment at CONUS scale for Topmodel and LASAM.
Steps to replicate behavior (include URLs)
1.
Screenshots