Scripts for calculating groundwater recharge via a stochastic vadose zone water budget

Description

This archive contains code related to the Open Source scientific computation software GNU Octave for a stochastic vadose zone water budget. A range of groundwater recharge estimates is generated using a Monte Carlo approach based on stochastic spatiotemporal rainfall variability and empirically-derived temporal evapotranspiration variability. This project is useful for calculating the uncertainty in recharge rates due to spatiotemporal variability in rainfall and annual variability in actual evapotranspiration. Including the spatial/temporal variability in rainfall reduces the uncertainty in precipitation and therefore uncertainty in recharge for a multi-decade water budget.

Steps to reproduce

Downloading and installing the Open Source scientific computation software GNU Octave (e.g., Eaton et al., 2021; https://octave.org/) will allow the user to run the scripts for an example dataset. The example is an updated version of Chapter 4 in Wiebe (2020). The README.txt file outlines the steps to prepare the input data and the order of the scripts. The user may run the scripts for other watersheds/datasets by compiling the list of input files outlined in the README.txt file. Example input data: The example input data were compiled from several sources. The evapotranspiration estimates were developed as described by Wiebe (2020) using daily temperature data from the Roseville weather station (Government of Canada, 2019), the ETo Calculator (Raes, 2009), and a local average wind speed estimate (Wiebe et al., 2019). The long-term rainfall and snowfall data were obtained for the Roseville weather station (Government of Canada, 2019) and missing values were filled in as noted in the README.txt file. The short-term rainfall data for the correlation analysis combined the Roseville data with data from the ~3-year Southern Ontario Water Consortium study (Wiebe et al., 2019). The actual evapotranspiration (AET) data (ratios of AET / precipitation) were compiled from the U.S. MOPEX dataset (Duan et al., 2006; Troch et al., 2018), where watersheds with long-term average potential evapotranspiration/precipitation (PET/P) ratios "similar to" (within +/-0.05 of) the ratio derived for the Roseville weather station were selected after screening out watersheds with poor quality based on the approach by Gentine et al. (2019) and a script shared by Pierre Gentine. The long-term total streamflow data were obtained for the New Dundee gauge operated by the Water Survey of Canada (WSC, 2019). Because the gauge is not located at the watershed outlet, a scaling factor was developed using the approach by Wiebe (2020) - Appendix Q - with updated numbers based on the datasets discussed here (e.g., fill in of missing precipitation values). Annual baseflow index (BFI) values were estimated using the Water Survey of Canada streamflow data based on the PART method within the U.S. Geological Survey Groundwater Toolbox software (Barlow et al., 2014). A correction factor to estimate AET from the vadose zone from total AET was derived using the approach in Appendix P of Wiebe (2020). References: Please see the "README.txt" file and the "Related links" section below.

Institutions

Categories

Funding

Related Links

Licence