Skip to content

Latest commit

 

History

History
46 lines (40 loc) · 3.02 KB

README.md

File metadata and controls

46 lines (40 loc) · 3.02 KB
Raw

Watershed Modelling Exercise 2

A repository for a conceptual hydrological model. (Exercise 2 Watershed Modelling, ETH Zurich)

Instructions set up Environemnt

  1. Download the zip file wat_bal_python.zip and extract it.

  2. Install python 3.7.6 (Required)Its Free!

    • To check that python is installed on your desktop:
      • For Mac open terminal and type $ python3 (Mac comes with preinstalled python 2.7.6 but you still need to install python 3.7.6)
      • For Windows 10, open Windows Power Shell and then type Python

  3. Install Visual Studio Code (Recommended)Its Free!

    • Visual Studio Code is an editor and will comes with autocomplete and syntax highlighting features
    • Any other editor should work as well but from here on the setup will be in Visual Studio Code
    • Download the version depending on your OS.

  4. Setting up the project. Installing packages.

    • Open Terminal/Windows Power Shell
    • Install pipenv by running $ pip3 install pipenv in Terminal/Windows Power Shell
    • Open Visual Studio Code
    • A welcome screen will appear, in the left side bar select the Extension (icon with 4 squares)
    • Search Python. Select Python (Linting, Debugging,..)
    • Once it is installed open Explorer on the top left bar. (shift+ctrl+e)
    • Select Open Folder, Browse to the directory water_bal_python.
    • The folder should contain Pipfile and Pipfile.lock
    • Go to Terminal at the top and select new terminal
    • Terminal will open at the bottom, Run the command $ pipenv install
    • This step may take some time and the internet connection is required
    • Restart the VSCode. Now on the lef side of blue colored bottom bar you can see the python version adn in bracket name of the folder along with pipenv. ('wat_bal_python:pipenv')

Instruction to run the scripts

  • There are 5 scripts namely model.py, plots.py, utils.py, run_opt.py, run_param_plot.py, run_sim.py and two folders name data and plots
  • All the required data is in the data folder
  • All the plots will be created in the plots folder
  • Two more files will be added namely parameter.h5 and gof.csv
  • Parameter.h5 will have the list of parameters for different runs
  • gof.csv will contain the goodness of fit values for observation and simulation
  • Run the scripts in order
    • run_opt.py will optimize the model and save the value of parameters in Parameters.h5
      • Number of runs nruns needs to be selected here. Roughly it takes 70-120 runs per iteration depending on the computer.
    • run_param_plot.py will generate the plots of the parameters
    • Seeing the histogram of parameters chose the value of the parameter and add the values to the script run_sim.py
    • Run run_sim.py to save the plots and gof values
    • Any change in the structure of the model needs to be made in model.py
    • For adding parameters to calibrate some changes also needs to be made in run_opt.py and run_sim.py