Deficit Irrigation with Aquacrop-OSPY

This report explores the efficacy of using Aquacrop OSPY to evaluate different irrigation schedules under deficit conditions. Aquacrop-OSPY is a Python implementation of an open source version of the Aquacrop model developped by the Food and Agricultural Organization. The introduction provides an overview of the climate and agricultural production of Tunisia, reviews the development of Aquacrop, Aquacrop-OS and Aquacrop-OSPY, and introduces the concept of deficit irrigation. The Aquacrop-OSPY module is then used to generate 1500 runs of the model (500 baseline runs and 1000 testing runs) varying irrigation dates and irrigation quantitaties. The data are visually inspected and an initial signifance test is evaluated.

See the Deployment here

---

Data

The data generated by the model runs can be accessed under the /model_results foder. The figures used in the report (and more) can be foud in the /figures folder. The code used to generate the model runs is in the /notebook folder. The notebooks which import this data and generate the charts and anylsis are also available in the /notebook folder.

---

Setup

Installing AquaCrop locally is tricky due to dependency issues. Here's the best way to get started:

For installation of the Aquacrop package please follow the instructions provided by Aquacrop-OSPY at the repository here. The notebooks require a few additional packages, for statistics and for plotting.

1. Use Python 3.8 or 3.9

• Version Compatibility: Make sure to use Python 3.8 or 3.9.
• Dependencies: Pay close attention to numba and numpy versions in error messages and install the required versions accordingly.

2. Use Google Colab

• Easiest Setup: Skip the local installation headaches by using the AquaCrop Google Colab environment set up by the AquaCrop team.
• How to Use:
  1. Open the Colab link.
  2. Import your notebook cells.
  3. Run the model to generate results.

For Plotting and Analysis

• Libraries Needed: seaborn and numpy.
• Steps:
  1. Install seaborn and numpy if not already available.
  2. Use these libraries to recreate charts and analyze data from the report.

---

Thanks

Thank you to Thomas Kelly for making the package available to Dr. Annelie Holzkämper for assigning such an interesting task for the Climate and Agiculture course at the university of Bern.