Gamifying weather forecasting: “Let it rain” campaign
This Inspire Challenge proposal was selected as a 2019 winner, with the team receiving 100,000 USD to put their ideas into practice.
The “Let it rain” campaign is a platform that will gamify weather prediction to incentivize farmers uptake of localised agro-advisories and help crowdsource weather information, which, when run through machine learning, will further improve weather forecasts.
Using Mediae’s popular farm make-over TV show Shamba Shape Up, the campaign will encourage farmers to guess the start of the rains, stirring up a national discussion on the relevance of weather forecasts and helping build farmer profiles for later providing customised climate info.
More about the project
The changing climate in East Africa coupled with the lack of timely and reliable agricultural information reaching farmers has led to inappropriate planting dates, which can lead to massive losses that threaten farmers’ livelihoods. A ten-day planting delay can reduce yield by up to 50 percent.
Even with an increasing number of organisations offering climate information services, only a small portion of farmers receive large-scale provision of big data-driven agronomic advice.
To address this issue, the project proposes to initiate innovative incentive mechanisms for the adoption of agro-weather advisories and crowd-sourcing using the Shamba Shape Up and iShamba platforms.
To take part in the “Let it rain” contest, farmers will need to subscribe to iShamba, becoming free members through the process, with the option of joining the premium service (600 KSh per year). The free service includes weekly information for two commodities of choice, two market prices of choice, and a weekly weather forecast. All iShamba members will have access to the call centre seven days a week and can send SMS questions that are answered by trained agronomists within 24 hours.
Reviewing the pool of farmers who submitted their predictions on the start of the rains through the “Let in rain” contest, and drawing from iShamba’s database of 275,000 users, the project will then select three groups of 1,500 farmers in one county, and ask them whether they would like to participate in a study on potato growing. The study will focus on the best planting date in relation to the start of the rains and other agronomic management activities that help farmers adapt to climate change in potato productions. Each of the three groups will receive different levels of information.
As the ability to predict the start of the rain improves, the data will be made available on Shamba Shape Up and iShamba to advise on planting dates and other agronomic information for crops. In order to provide weather forecast and optimal planting date advisory, the Alliance of Bioversity International and CIAT will work with Kenya Agricultural and Livestock Research Organization’s (KALRO) Big Data Weather Platform and the International Potato Center (CIP).
Alliance of Bioversity International and CIAT
Step by step
Assessing COVID-19 impacts
In 2021, the team is pivoting its activities towards assessing the COVID-19 impacts on farming and understanding how farmers across Kenya have adjusted to lockdown measures during the pandemic.
Learn more about the team’s Inspire Challenge Rapid Response Grant work here.
In collaboration with Usiku Games, a gaming studio based in Nairobi, a game has been developed to facilitate a two-way conversation with farmers about changing weather patterns, the effects, and mitigation strategies.
Farmers will receive a SMS prompting them to input their guesses and responses to follow-up questions via a chatbot created by the team.
1,000 USD of prize money will be split among the winner(s) in each county.
Let the games begin
The game was launched during the second weekend of February 2020, in tandem with this season’s launch of the Shamba Shape Up TV program.
The game was advertised via video ads (skits) and promoted on the online iShamba platform, as well as through the Mediae Company Twitter and Facebook.
Game registration and playing period
25,000 users signed up to play the “Let it rain” game via the iShamba platform. Participants submitted their guesses for when the rains would start in their county.
The game was closed in mid-March, four weeks after opening.
Tracking the rains
Team members from the Alliance of Bioversity International and CIAT tracked the first rainfalls in each of the 10 counties. Then, the rainfall dates were compared to participants’ guesses to see who guessed the rains correctly.
Winners announced and prizes awarded
245 winners were announced across 10 counties for their correct guesses for the date of the first rains. Winners received their prize money (1,000 USD split among the winner(s) in each county) via M-Pesa, a mobile phone-based money transfer service.
The prize money was distributed in the first months of the COVID-19 pandemic, during a period of decreased income for many, and it was found that winners commonly invested this money back in their farms. Farm input purchases and small livestock investments were the most common uses of the funds.
Watch the video below to hear from a winner in Kiambu county:
Project awarded US$40K CGIAR Rapid Response grant
In response to the food security issues brought about by the COVID-19 pandemic, CGIAR awarded the team a Rapid Response grant to develop a mobile phone-based panel survey approach to assess COVID-19 impacts on farming, which will be tested at scale across all counties in Kenya through the “Let it rain” campaign.
The findings of this survey will inform farmers with targeted advisories through the Mediae digital media platforms, and inform programmatic responses through info briefs.
Learn more about the project, Rapid diagnostics of COVID-19 farming impacts.
Project awarded US$100K scale-up grant
The project was one of four winners of the Inspire Challenge 2019 and was awarded US$100K at the Convention of the CGIAR Platform for Big Data in Agriculture, during 16-18 October, 2019.
Selection of counties
Ten counties in Kenya were selected as the project’s focus region. The counties were chosen to fairly represent various regions.
Weather data analysis
The CIAT team has begun analysis of 10 years of weather data (from aWhere, facilitated by shared services provided by the BIG DATA Platform), to see what the algorithms would say is the historical start of the rainy season.
25 ml of rain would be needed in one day to trigger the start of the rainy season for the algorithm.
Based on the analysis, the latest onset in any of the ten included counties would occur in mid-April, and central counties would experience onset in late-March.
Gender & Youth Inclusion
- The team collected demographic data from “Let it rain” campaign participants when registering on the iShamba platform, which could inform a gender analysis of the rain guess data.
Project News and Resources
Meet all the Winners