Running Drought Indices from Jupyter Notebooks

Here is a simple workflow using examples, locations and dates that will aid you in accessing the ANIN drought indices through Jupyter Notebook. After completing the steps below, you should be able to adjust the location with your coordinates and date ranges where applicable.

The steps described below should be completed after the initial Installation.

Install Jupyterlab and Notebook

Use the following Python prompts to install Jupyterlab and Notebook.

pip install jupyterlab

pip install notebook

Launch Jupyter Notebook

1. Open Windows Explorer and navigate to the folder where the ANIN-drought-indices folder is cloned.
2. Click on the Address bar and copy the address.
3. Type CMD in the address bar and hit Enter. This will launch the Command Prompt.
4. In the Command Prompt window, enter the following code:

   jupyter notebook
   

   This will open Jupyter Notebook in your default browser.

Setup and Testing

Double-click setup_and_test.ipynb on the home page. It will open in a new tab and require you to 'Restart the kernel and run all cells (under Run in the menu). Jupyter Notebook will install all the requirements to run the code, followed by a sample test query, which requests a Sentinel-2 image over The Innovation Hub in Pretoria, South Africa.

Edit the spatial and temporal extent

Within the code block, you can change the location to that of your choosing and adjust the date range for the image search and the name of the output file by changing the values in the highlighted lines shown below.

import openeo
from openeo.processes import process

url = "https://openeo.cloud"
connection = openeo.connect(url).authenticate_oidc()

spatial_extent_sansa = {
    "east": 28.275,
    "north": -25.740,
    "south": -25.755,
    "west": 28.260,
}

datacube = connection.load_collection(
    collection_id="SENTINEL2_L2A",
    bands=["B04", "B03", "B02"],
    spatial_extent=spatial_extent_sansa,
    temporal_extent=["2024-03-05", "2024-03-07"],
)
# Scale values for better visualisation
datacube = datacube / 2500

datacube.download("out-Pretoria.tiff")

Run Standardised Precipitation Index (SPI)

In the SPI folder, you will find SPI_openeo.ipynb. Double-click on it to open it in a new tab. This index will return results for the whole country with results previews provided.

The code block provides you the option to adjust the temporal extent by adjusting the highlighted line of code:

dc = SPI_dc
dc = dc.filter_temporal(["2022-01-01", "2022-06-01"])
# out_format = "NetCDF"
out_format = "GTiff"
dc = dc.save_result(format=out_format)
output_path = custom_execute_batch(dc, job_options=heavy_job_options, out_format=out_format)

Run Soil Moisture Anomaly (SMA)

From the Jupyter Home page, open the SMA folder and double-click SMA_openeo.ipynb to open the notebook. It will open in another tab. As will the startup, Restart the kernel and run all cells. Should you encounter a request to authenticate your account, follow the link to the EGI login website and log in. Return to the SMA notebook to view the results.

Tip

The script will create the output TIFFs in folders labelled out-xxx in the respective folders, which can be visualised in any GIS software.