Exercise: exploring drought in Kenya

We will now move on to use the CATE toolbox from within Jupyter notebook to retrieve and analyse climate records of soil moisture. After doing this exercise, you will have a good idea about soil moisture, an essential climate variable, and how to analyse it to identify and quantify drought events.

Introduction

Kenya is a drought-prone country (see Figure 1), mainly due to its particular environmental conditions. Agriculture supports up to 75% of Kenya’s population and generates almost all food requirements and drought is a significant constraint on rain-fed agricultural production, particularly in Kenya’s arid and semi-arid parts. In the last century, Kenya has faced almost 29 droughts, some of them in the past decade. Droughts in Kenya appear to be becoming more frequent and severe over time. As a result, crop failures and livestock deaths lead to severe food shortages in Kenya and food insecurity. The Climate Change Initiative, CCI hosts a soil moisture dataset that can generate an index of drought severity. In this exercise, you will be guided to retrieve Soil Moisture Deficit Index (SMDI) from soil moisture data generated by the Soil Moisture-CCI project.

Figure 1: Drought leaves dead and dying animals in northern Kenya. Oxfam International, licensed under CC BY-NC-ND 2.0. (Click to expand)

This exercise explains how to identify the drought period in Kenya using soil moisture data from the open data portal to create a drought indicator. This exercise is based upon the Jupyter notebook. You are guided to use the Jupyter notebook and Climate Analysis Toolbox of the ESA CCI – Cate to apply soil moisture products provided by the soil moisture-CCI project.

Learning outcomes

1- Identify and access and soil moisture data from the Open Data Portal.

2- Create a drought index based on the soil moisture dataset for Kenya.

3- Identify periods and regions of drought in Kenya.

The input data is daily images of volumetric soil moisture content (SM, m³/m³) for Kenya and several years (2015-2020). There are different approaches to obtaining periods of droughts from soil moisture information. In this exercise, you will follow the ‘Soil Moisture Deficit Index’ to address drought in Kenya.

Preparations

If you haven’t done so please follow the CATE from Jupyter notebooks to get started using the Jupyter notebook exercises.

CATE software and the open data portal

We’ll be querying the open data portal for soil moisture products produced by the Climate Change Initiative Soil Moisture project. The open data portal hosts a variety of datasets on the Jasmin infrastructure in the UK. These can be queried using the dedicated CATE python module. CATE allows retrieving data from the data portal using its web-interface.

In this Jupyter notebook exercise, we will directly call CATE functions and show you how these can be used in advanced workflows.

Starting the exercise

To perform the exercise and execute the code in this notebook you can run it through this link: Click to run the exercise.

Alternatively, you can download/clone the notebooks from github and run them through your own Jupyter notebook instance (this may initially take more time to setup, but does allow you to store your changes and extend the notebook exercises to your own needs).

Up next

Once you have done the exercise, it’s time to discuss what you can learn from the data.