The Database That’s Fighting Drought

One computer scientist is combining traditional farming techniques with weather data to help farming families in Kenya.

Long before Prof. Muthoni Masinde became an Associate Professor and Head of the IT Department at Central University of Technology, Free State, South Africa, she lived in a remote village in the eastern part of Kenya. The Mbeere area is home to many farming families who grow maize, cowpeas, millet, and sorghum. Muthoni’s relatives still live in the area and she visits often, taking time to talk with her mum and farming neighbors.

Over the years, Muthoni has witnessed how a few raindrops can make or break livelihoods. Droughts can easily destroy farmers’ hard work in a place where crops rely only on rain, instead of irrigation, for survival. And in sub-Saharan Africa, particularly the Greater Horn of Africa, droughts are an increasingly urgent problem.

According to Oxfam, rampant droughts in the region have left millions of families hungry and not knowing where their next meal will come from. Unpredictable weather patterns and sporadic rainfall mean farmers can’t accurately plan their crops. As a result, crops that could feed hungry families often fail and die. “The problem is only becoming worse,” explains Muthoni, “This year will be 30 percent worse than last year.”

Farmers aren’t just losing crops, they are beginning to lose hope.

Although meteorological organizations in these countries provide regular seasonal forecasts, the local farmers have trouble accessing that information, especially in rural villages with little to no access to television, newspapers, and other news sources. And even when they can access these forecasts, farmers don’t know how to interpret figures, such as average temperature and near-normal rainfall.

Instead, these farmers rely on their traditional methods and describe the weather in tangible ways — such as “the smaller variety of mango tree has produced lots of fruit this season,” meaning that it is a poor rainy season. Or “the migratory birds are arriving in large numbers.” This means rain will fall in a week’s time, so farmers know they need to complete planting.

Small-scale farmers in villages like the one Muthoni grew up in don’t stand to lose just crops, they are also beginning to lose hope. Many are so discouraged by the droughts that they give up farming — further threatening national food supplies.

“At the end of the day, the farmers just want to produce enough food to sustain themselves and their families,” says Muthoni, “If they can do that, they’ll be happy.”

Bringing science and indigenous knowledge together

I knew I could package indigenous knowledge so it could be used to help farmers easily predict droughts.

As a computer scientist who grew up among farmers, Muthoni was in a unique position to create an innovative tool that farmers in Mbeere and elsewhere could use to predict and plan for droughts.

Muthoni knew that most farmers use indigenous knowledge to predict the weather — indicators like the blooming of trees or the movement of animals can let them know if rainfall is approaching. However, that information alone didn’t allow them to predict the weather beyond a single season.

Muthoni realized that the key to an accurate and usable tool was combining data from conventional weather stations and wireless sensors with farmers’ traditional knowledge. She has created an integrated weather information system with this data. Using Artificial Neural Networks, a computer model based on the structure of biological neural networks, she has built computer models capable of predicting droughts up to four years in advance.

That information led to the development of Muthoni’s Drought Prediction Tool called ITIKI (Information Technology and Indigenous Knowledge with Intelligence). ITIKI informs farmers through smartphone alerts of upcoming droughts, using the language they understand.

For example, farmers could receive a message alerting them that rain will be less and later than usual — suggesting they should plant a crop that requires less water. Or that dragonflies have been spotted flying close to the ground, meaning it will be raining in 3 days’ time. Muthoni is also providing the farmers with smartphones that quickly receive the alerts, allowing them time to plan their planting and harvesting with precision.

“I knew I could package indigenous knowledge so it could be used to help farmers easily predict droughts,” she says.

As she continues to develop and update ITIKI, Muthoni speaks to the farmers constantly. They are eager to teach others about their indigenous methods and knowledge. “Their faces glow because they realize that their knowledge is useful,” she says.

A Bright Future for Farmers

Muthoni explains that while ITIKI has been tested with only 200 farmers in three locations, she’s received a grant from USAID’s Securing Water for Food (SWFF) and plans to expand it to up to 4,000 farmers in more locations in three counties.

“We want farmers all over to know about this tool,” Muthoni says. “Lives will be changed if farmers use this.”

USAID, Sweden through the Swedish International Development Cooperation Agency (Sida), the Government of South Africa, and the Ministry of Foreign Affairs of the Kingdom of the Netherlands provide this innovator with funding and technical assistance.