header image

Snow Cover Forecast Model Tutorial

Snow Cover Forecast Model Tutorial

The Snow Cover Forecast Model is designed to function entirely in a web-browser using Google Sheets, a web-based version of a spreadsheet. A description of model’s inner workings can be found here.

To test the model for yourself use the editable version set up for the La Laguna watershed in central northern Chile. Follow these steps and recommendations:

  1. Use Google Chrome as your web-browser. This is not an endorsement of Chrome, but Sheets and its add-ons simply work better in the Google system.
  2. Remove the data values for cells B1 – B4. These are the four parameters needed for the model. Removing them will provide you the opportunity to calibrate the model.
  3. To calibrate the model you must install the Add-on  “SolverSolver is a tool that finds the best solutions for the four parameters, optimizing model performance. The Solver plugin works well in Chrome, but not in other browsers.
  4. Once Solver is available as an Add-on, you should activate it.
  5.  A dialogue box should open on the right of your screen. Set the dialogue box to the following:
    1. Set Objective – E1 (This is the Nash-Sutcliffe parameter, which is a measure of model performance)
    2. Select Max button in order to optimize model performance.
    3. By changing – B1:B4
    4. For Solving Method : LSGRG Nonlinear*
    5. Choose Solve

Want a visual example? Please refer to the following screenshot.

Once you have run Solver the four parameter values (B1 – B4) will populate, optimize model performance, and calculate the Nash-Sutcliffe values (Cell E1).

Additionally you have the option of setting a minimum Nash-Sutcliffe value in cell H1.

There is also a second tab in the file that shows the modeled results.

A validated, functioning model for the La Laguna watershed in northern central Chile can be found at https://goo.gl/4QNQL3.

There are also models set up for the Rio Aragon watershed (goo.gl/6fYGTo) in northern Spain and the John Day watershed (goo.gl/CP5Ev1) in central Oregon, USA

Thanks for your time, expertise and input! Now please head back to the survey to give us your feedback.

http://tiny.cc/snowEv

Thanks again – you rock!

If you have any questions regarding the model, or suggestions of how this tutorial can be improved, please contact sprolese <at> oregonstate.edu.
https://www.linkedin.com/in/eric-sproles/

* LSGRG stands for Least Squares Generalized Reduction Gradient

Print Friendly, PDF & Email