Skip to content

Latest commit

 

History

History
 
 

bmi_config_files

README.md

BMI Configuration

BMI requires a configuration file for each model. The LSTM configuration files contains key value pairs that are used by the BMI to run the model. Below are examples and descriptions of each of such keys, and what type of values are associated.

Initialization Information

These key value pairs are used by the BMI to set up the model in some particular way

  • train_cfg_file: ./trained_neuralhydrology_models/hourly_all_attributes_and_forcings/config.yml found here. This is a very important part of the LSTM model. This is a configuration file used when training the model. It has critical information on the LSTM architecture and should not be altered.
  • initial_state: 'zero' This is an option to set the initial states of the model to zero.
  • verbose: 0 Change to 1 in order to print additional BMI information during runtime.

Static Attributes

These are static attributes that are particular to the catchment. These should be calculated in the same manner as the values which the LSTM was trained. Some description is provided below, but again see Addor et al. 2017 for more details.

  • area_sqkm: 620.38 allows bmi to adjust a weighted output
  • elev_mean: 92.68 catchment mean elevation (m) above sea level
  • slope_mean: 17.79072 catchment mean slope (m km−1)
  • area_gages2: 573.60000 catchment area (GAGESII estimate), (km2)
  • frac_forest: 0.9232 forest fraction
  • lai_max: 4.87139 maximum monthly mean of the leaf area index (based on 12 monthly means)
  • lai_diff: 3.74669 difference between the maximum and minimum monthly mean of the leaf area index (based on 12 monthly means)
  • gvf_max: 0.863936 maximum monthly mean of the green vegetation fraction (based on 12 monthly means)
  • gvf_diff: 0.337712 difference between the maximum and minimum monthly mean of the green vegetation fraction (based on 12 monthly means)
  • soil_depth_pelletier: 17.412808 depth to bedrock (maximum 50 m) (m)
  • soil_depth_statsgo: 1.491846 soil depth (maximum 1.5 m; layers marked as water and bedrock were excluded) (m)
  • soil_porosity: 0.415905 volumetric porosity (saturated volumetric water content estimated using a multiple linear regression based on sand and clay fraction for the layers marked as USDA soil texture class and a default value (0.9) for layers marked as organic material; layers marked as water, bedrock, and “other” were excluded)
  • soil_conductivity: 2.375005 saturated hydraulic conductivity (estimated using a multiple linear regression based on sand and clay fraction for the layers marked as USDA soil texture class and a default value (36 cm h−1) for layers marked as organic material; layers marked as water, bedrock, and “other” were excluded) (cm h-1)
  • max_water_content: 0.626229 maximum water content (combination of porosity and soil depth statsgo; layers marked as water, bedrock, and “other” were excluded)
  • sand_frac: 59.390156 sand fraction (of the soil material smaller than 2mm; layers marked as organic material, water, bedrock, and “other” were excluded)
  • silt_frac: 28.080937 silt fraction (of the soil material smaller than 2mm; layers marked as organic material, water, bedrock, and “other” were excluded)
  • clay_frac: 12.037646 clay fraction (of the soil material smaller than 2mm; layers marked as organic material, water, bedrock, and “other” were excluded)
  • carbonate_rocks_frac: 0 fraction of the catchment area characterized as “carbonate sedimentary rocks”. GLiM
  • geol_permeability: -14.2138
  • p_mean: 3.60813 mean daily precipitation (mm day-1)
  • pet_mean: 2.11926 mean daily PET, estimated by N15 using Priestley–Taylor formulation calibrated for each catchment (mm day-1)
  • aridity: 0.587356 aridity (PET /P, ratio of mean PET, estimated by N15 using Priestley–Taylor formulation calibrated for each catchment, to mean precipitation)
  • frac_snow: 0.245259 fraction of precipitation falling as snow (i.e., on days colder than 0 C)
  • high_prec_freq: 20.55 frequency of high precipitation days (≥5 times mean daily precipitation) (days yr-1)
  • high_prec_dur: 1.20528 average duration of high precipitation events (number of consecutive days ≥5 times mean daily precipitation)
  • low_prec_freq: 233.65 frequency of dry days (< 1mmday−1) (days yr-1)
  • low_prec_dur: 3.66223 average duration of dry periods (number of consecutive days < 1mmday−1) (days)

Optional Metadata

These key value pairs contain metadata that are not required to run the model but can be useful to make sure that the model is running as expected. It is best to consider items listed here as optional, but also as potential enhanced development looking ahead.

  • time_step: '1 hour' As of this writing, both time_step_size and time_units are defined during bmi.initialze(). The next phase of development will provide the model with all time information via bmi configuration.
  • basin_name: 'Narraguagus River at Cherryfield, Maine' Not currently directly used but may be beneficial for bookkeeping.
  • basin_id: '01022500' Future development will require unique ID for node-to-node routing; still under beta
  • lat: 44.60797 Post-run analysis or plotting only
  • lon: -67.93524 Post-run analysis or plotting only