What is it about?

The East African ‘short rains’ usually occur in the period between October and December, but this varies greatly from year to year. If the rains fail, droughts can occur, especially when several consecutive seasons fail, and if it rains abundantly, there is a risk of flooding. To predict the rainy seasons, statistical models, which make use of a relationship between ocean temperatures in the Indian Ocean and the Pacific Ocean, are often used. We make a simple statistical model and study the cases when the model fails to predict the short rains.

Featured Image

Why is it important?

As statistical models to predict rainfall are widely used in East Africa, it is important to study the performance of such models, so that we can learn more about the linkages between ocean temperatures and rainfall, as well as assess the years when statistical forecasts fail. We checked if upfront adjustments of the model might be useful and found that it would have helped during the dry 1987 season, for which the model erroneously predicted copious rainfall, but it would have exacerbated the forecast for the wet 2019 season, when the model predicted too-dry conditions.


This was my third paper on African climate, and I found it very interesting to look for systematic errors in the prediction models. Through my work with climate institutes in East Africa, such as ICPAC and national meteorological services, I have a chance to influence the way forecasts are created and used, and I hope the new knowledge presented in our paper will be used in the interpretation of seasonal rainfall forecasts.

Dr Erik W Kolstad
NORCE Norwegian Research Centre

Read the Original

This page is a summary of: Lagged oceanic effects on the East African short rains, Climate Dynamics, February 2022, Springer Science + Business Media, DOI: 10.1007/s00382-022-06176-6.
You can read the full text:



The following have contributed to this page