SEN4RUST
HOW EARTH OBSERVATION IS HELPING ETHIOPIA’S FARMERS TO THRIVE
Wheat is deeply woven into Ethiopia’s history and culture. Yet as destructive crop diseases threaten the livelihoods of millions of farmers, scientists are developing and improving a new early-warning system to prevent outbreaks and turn the tide.
For Ethiopia’s wheat farmers, 2010 was supposed to be a good year.
For Ethiopia’s wheat farmers, 2010 was supposed to be a good year.
But as the growing season was underway, a devastating plant disease took hold among the crops. Local wheat varieties had no defence against the new pathogen, a strain of yellow rust which had blown over from the Himalayas. Weather conditions helped the disease spread quickly. The resulting outbreak affected around 600,000 hectares of wheat, causing economic losses upwards of $250 million USD. It was a stark reminder of the constant battle Ethiopia’s farmers face with plant disease, and a setback in the country’s long journey toward becoming a self-sufficient producer of wheat.
Ethiopia’s farmers have cultivated local forms of durum wheat for thousands of years. In the 1940’s, non-Ethiopian bread wheat was brought in with the hope of boosting yields. Yet due to a combination of low yield varieties, limited inputs and poor management techniques, production levels remained stagnant for decades.
Thanks to the widespread expansion of irrigation systems, the area of land dedicated to wheat production has grown from around 1.5 million hectares in 2010 to 2.5 million hectares in 2023. Ethiopia is now the largest wheat producer in sub–Saharan Africa by a long way.
Most of this production is done by the 4-5 million smallholder farm households dependent on wheat for food and income.
Wheat is ground into flour and used to prepare bread, pasta and traditional foods or is roasted or boiled and eaten as grain. Straw from the wheat is also a valuable source of animal feed, and is used in thatched roofing.
This humble grain is therefore enormously important for food security, but economically and politically too.
Unfortunately, the widespread expansion of production across the country has led to a significant uptick in plant disease.
This includes wheat rusts, one of the most significant biological threats to the crop. These destructive fungal pathogens spread quickly and can devastate crop yields. Wheat yellow rust, a major fungal disease, is spreading around the world — causing crop losses of up to 5.5 million tons per year.
It can be tricky to measure the exact impact of diseases on crop yields however, partly because a good year for disease — adequate rain and sunlight — is also a good year for crops. This causes a ‘masking effect’ in the data, hiding the true extent of crop losses due to disease.
In Ethiopia, there are clearly visible production dips in 2010 and 2016 and wheat rust outbreaks were likely contributory factors. The rust outbreak in 2010 spread to over a third of Ethiopia’s wheat production areas, and destroyed around 20% of that year’s harvest.
As wheat consumption in Ethiopia still outstrips production, poor harvests can have widespread consequences. Ethiopia’s farmers already faced global challenges such as price fluctuations and climate change.
Now these diseases increasingly
threaten their wheat.
Unfortunately, the widespread expansion of production across the country has led to a significant uptick in plant disease.
This includes wheat rusts, one of the most significant biological threats to the crop. These destructive fungal pathogens spread quickly and can devastate crop yields. Wheat yellow rust, a major fungal disease, is spreading around the world — causing crop losses of up to 5.5 million tons per year.
It can be tricky to measure the exact impact of diseases on crop yields however, partly because a good year for disease — adequate rain and sunlight — is also a good year for crops. This causes a ‘masking effect’ in the data, hiding the true extent of crop losses due to disease.
In Ethiopia, there are clearly visible production dips in 2010 and 2016 and wheat rust outbreaks were likely contributory factors. The rust outbreak in 2010 spread to over a third of Ethiopia’s wheat production areas, and destroyed around 20% of that year’s harvest.
As wheat consumption in Ethiopia still outstrips production, poor harvests can have widespread consequences. Ethiopia’s farmers already faced global challenges such as price fluctuations and climate change.
Now these diseases increasingly
threaten their wheat.
THE GLOBALISATION OF PLANT DISEASE
Around the world, pathogens and the diseases they cause are on the rise. Agricultural crop production is under threat, particularly in Sub-Saharan Africa.
Climate change is altering weather patterns, environments and ecosystems, which can allow plant diseases to thrive in new areas.
Our planet is also increasingly linked through complex networks of transport and trade, which facilitate the spread of disease.
And pathogens can spread with wind systems, travelling vast distances across borders. Rusts can travel tens of thousands of kilometres on the wind, crossing countries and even continents.
The world is seeing increasingly frequent and severe crop disease outbreaks, which are also causing more virulent strains to emerge.
The more disease that exists, the higher the likelihood that new strains will evolve, with random mutations able to overcome previous resistance in crops. This was the case in Ethiopia’s major wheat outbreak in 2010.
Ethiopia’s location at the centre of several major crop production regions also makes it more likely that pathogens will make their way in or out.
Several major airborne dispersal routes pass into and out of East Africa, toward Central and South Asia via the Middle East.
This makes Ethiopia a priority country for the surveillance and monitoring of wheat rust.
Against this backdrop, scientists are developing early warning systems to help prevent the spread of disease.
The world is seeing increasingly frequent and severe crop disease outbreaks, which are also causing more virulent strains to emerge.
The more disease that exists, the higher the likelihood that new strains will evolve, with random mutations able to overcome previous resistance in crops. This was the case in Ethiopia’s major wheat outbreak in 2010.
Ethiopia’s location at the centre of several major crop production regions also makes it more likely that pathogens will make their way in or out.
Several major airborne dispersal routes pass into and out of East Africa, toward Central and South Asia via the Middle East.
This makes Ethiopia a priority country for the surveillance and monitoring of wheat rust.
Against this backdrop, scientists are developing early warning systems to help prevent the spread of disease.
How Wheat Rust infects Crops
First, the spores land on the crop and penetrate into green parts of the plant seen during early growth.
Fungal fibres known as hyphae then spread and grow within the plant.
Depending on weather conditions, the disease will take around two weeks to fully develop. At this point, stripes or pustules appear on the leaves or stems.
First, the spores land on the crop and penetrate into green parts of the plant seen during early growth.
Fungal fibres known as hyphae then spread and grow within the plant.
Depending on weather conditions, the disease will take around two weeks to fully develop. At this point, stripes or pustules appear on the leaves or stems.
Once the spores become visible, they are released into the wind to infect other plants. The cycle continues.
Each crop can produce billions of spores.
The combination of susceptible crops and favourable weather conditions creates the potential for huge, explosive epidemics over vast areas.
Once the spores become visible, they are released into the wind to infect other plants. The cycle continues.
Each crop can produce billions of spores.
The combination of susceptible crops and favourable weather conditions creates the potential for huge, explosive epidemics over vast areas.
Ethiopia is especially sensitive to wheat rust because of its unique environment and cropping patterns.
With a diverse agro-ecological landscape and variable conditions, farmers are able to plant crops in two wheat-growing seasons, known as Meher (main) and Belg (short rains).
Nearly 90% of wheat is grown across the country during the main growing season, Meher, which starts around June and lasts until December.
In some areas, particularly in the south-eastern highlands, farmers also grow wheat in a shorter season, Belg. Planting begins around March and crops are harvested in July.
Ethiopia is especially sensitive to wheat rust because of its unique environment and cropping patterns.
With a diverse agro-ecological landscape and variable conditions, farmers are able to plant crops in two wheat-growing seasons, known as Meher (main) and Belg (short rains).
Nearly 90% of wheat is grown across the country during the main growing season, Meher, which starts around June and lasts until December.
In some areas, particularly in the south-eastern highlands, farmers also grow wheat in a shorter season, Belg. Planting begins around March and crops are harvested in July.
Ethiopia is especially sensitive to wheat rust because of its unique environment and cropping patterns.
With a diverse agro-ecological landscape and variable conditions, farmers are able to plant crops in two wheat-growing seasons, known as Meher (main) and Belg (short rains).
Nearly 90% of wheat is grown across the country during the main growing season, Meher, which starts around June and lasts until December.
In some areas, particularly in the south-eastern highlands, farmers also grow wheat in a shorter season, Belg. Planting begins around March and crops are harvested in July.
TRACKING
PLANT DISEASE FROM SPACE
In SEN4RUST, a project supported by the European Space Agency and CIMMYT, researchers have tested the added-value of Earth observation data on an existing early warning system to counter the rising threat of wheat rusts and support farmers on the ground. It was established by a consortium of national and international partners: Cambridge University, UK Met Office, International Maize and Wheat Improvement Center (CIMMYT), Ethiopian Institute of Agricultural Research (EIAR), Ethiopian Agricultural Transformation Agency Institute (ATI) and Ethiopian Ministry of Agriculture.
The Ethiopian Wheat Rust Early Warning and Advisory System (EWAS) is one of the most advanced, operational national crop disease early warning and advisory systems in the world.
EWAS draws on an integrated system of ground surveillance, weather forecasts and satellite observations, along with atmospheric and epidemiological models.
On the ground, specialists survey wheat fields looking for rusts and taking samples to decipher which strain is present on the crop.
This information is collected on smartphones and automatically feeds into a shared database among the EWAS partners.
Using advanced forecast models, the team can predict the ways the fungal spores will disperse and where they will land. They can also detect where environmental conditions are ripe for rusts to take hold and develop.
EWAS draws on an integrated system of ground surveillance, weather forecasts and satellite observations, along with atmospheric and epidemiological models.
On the ground, specialists survey wheat fields looking for rusts and taking samples to decipher which strain is present on the crop.
This information is collected on smartphones and automatically feeds into a shared database among the EWAS partners.
Using advanced forecast models, the team can predict the ways the fungal spores will disperse and where they will land. They can also detect where environmental conditions are ripe for rusts to take hold and develop.
This information forms the basis of a detailed bulletin, which is distributed on a weekly basis to regional and local authorities across Ethiopia. Farmers are then quickly and directly informed of any potential outbreaks.
Ethiopia’s Agricultural Transformation Institute (ATI) has also set up a mobile phone voice alert, with over five million farmers subscribed throughout the country.
With the new surveillance system in place, farmers get around three weeks notice for any new potential outbreak. Since the disease takes roughly two weeks to develop in a new area, this gives them enough warning to start preparing and implementing control measures.
EWAS has already delivered tangible, sizable impacts by stopping a potentially destructive outbreak in its tracks.
In 2021, a new strain of wheat rust appeared in Ethiopia, threatening a production wipeout similar to 2010.
Another highly virulent strain had appeared with links back to the Himalaya region. Weather conditions were optimal, and farmers around the country were growing a lot of susceptible wheat varieties in their fields.
This time, with the early warning system in place, the outcome was very different. The situation flipped from a potentially disastrous epidemic to a bumper harvest. In the end, 2021 was a very good year for Ethiopian wheat.
Ethiopia’s farmers can now grow their crops, knowing their livelihoods are protected by eyes in the sky looking out for harmful plant disease. And Ethiopia’s dream of becoming a self-sufficient wheat producer may soon become a reality.
The Sen4Rust project has tested if integrating Earth Observation data in EWAS could provide an ever more detailed picture of the emergence and spread of disease.
Currently, a missing piece of the puzzle in EWAS is the up-to-date knowledge about the crop itself. Detailed information is lacking on where the wheat is grown in each season, on its phenology (seasonal timing) and on the dynamics of the wheat growing cycle. In Ethiopia, this information is critical due to its complex environmental conditions, and the green bridging effect between two growth seasons.
The objective of the Sen4Rust project was to derive this information by remote sensing — a way of monitoring plants growth from satellites through measurements of their reflected and emitted radiation.
Using images from Europe’s Sentinel-2 Satellite, the scientists can establish — in near-real time — exactly where and when wheat is being grown.
The green line shows Oromia, the main wheat growing region of Ethiopia. Each square represents one Sentinel-2 tile.
The objective of the Sen4Rust project was to derive this information by remote sensing — a way of monitoring plants growth from satellites through measurements of their reflected and emitted radiation.
Using images from Europe’s Sentinel-2 Satellite, the scientists can establish in near-real time exactly where, and when wheat is being grown.
The green line shows Oromia – the main wheat growing region of Ethiopia, each square representing one Sentinel-2 tile.
First, the distribution of wheat is mapped by combining ground truth crop surveys and high spatial resolution satellite imagery down to a scale of 10 meters. This map reveals the potential susceptibility of crops grown around the country to wheat rusts.
First, the distribution of wheat is mapped by combining ground truth crop surveys and high spatial resolution satellite imagery down to a scale of 10 meters. This map reveals the potential susceptibility of crops grown around the country to wheat rusts.
Sentinel-2 time series are also used to describe the wheat phenology (start, peak and end of the season). This feeds into high-resolution maps showing greenness changing through time, to reveal where plants could be hosting the disease.
If the forecasts indicate that spores will spread into other wheat-growing areas, where crop varieties are susceptible and green parts are readily available, the scientists can pinpoint areas of real risk.
This map combines EWAS grids with phenological data, showing the timing of wheat growth across the region.
Orange and red dots show unimodal growth where wheat is grown only during the main rain season between June and November
The blue dots indicate spots where growth is bimodal — when crops are planted during both Belg and Meher.
Green dots represent areas with long duration green vegetation, which are likely to include some wheat at low density.
With the added earth-observation data, the model can now reveal more detailed predictions of spore load, accounting for areas where different growth seasons overlap.
Even with Ethiopia’s complicated growing seasons and cropping patterns, scientists can now figure out where green crops will be available at any time of the year — and where the pathogen might thrive. The net impact is a more accurate prediction of wheat stem rust risk, leading to more informed and targeted early warning advice to smallholder farmers.
Orange and red dots show unimodal growth where wheat is grown only during the main rain season between June and November
The blue dots indicate spots where growth is bimodal — when crops are planted during both Belg and Meher.
Green dots represent areas with long duration green vegetation, which are likely to include some wheat at low density.
With the added earth-observation data, the model can now reveal more detailed predictions of spore load, accounting for areas where different growth seasons overlap.
Even with Ethiopia’s complicated growing seasons and cropping patterns, scientists can now figure out where green crops will be available at any time of the year — and where the pathogen might thrive.
The net impact is a more accurate prediction of wheat stem rust risk, leading to more informed and targeted early warning advice to smallholder farmers.
These new integrations have boosted the system’s accuracy, allowing it to rely on highly detailed observations of wheat down to the smallholder farm level.
The SEN4RUST team is now considering a similar system, which will be able to detect new irrigated areas across the country. This will add another layer of accuracy to the forecasts.
This image shows an irrigated area, with arid land surrounding green vegetation. The yellow shows the distribution of wheat, detected by the system’s algorithm.
These new integrations have boosted the system’s accuracy, allowing it to rely on highly detailed observations of wheat down to the smallholder farm level.
The SEN4RUST team is now considering a similar system, which will be able to detect new irrigated areas across the country. This will add another layer of accuracy to the forecasts.
This image shows an irrigated area, with arid land surrounding green vegetation. The yellow shows the distribution of wheat, detected by the system’s algorithm.
EWAS has expanded due to its success in Ethiopia, countering the spread of wheat rusts. New surveillance networks have already been set up in Nepal and Bangladesh. Other teams are developing similar systems in Kenya, Tanzania, Zambia, Pakistan, and Bhutan.
Millions of farmers across Africa and Asia now have access to early warning systems, helping them to fight the scourge of wheat rusts that has plagued the crop for thousands of years.
Other scientists are adapting the idea to tackle other plant diseases.
One major target is wheat blast, a devastating new crop disease that emerged in recent decades and is likely to expand as a result of climate change.
The Sen4Rust demonstration is the first successful integration of landscape level wheat vulnerability maps with advanced epidemiological models.
It represents a significant step change in the fight against crop disease, and could advance disease forecasting — as well as our knowledge of the underlying pathogens.
Following EWAS expansion, the integration of Earth observation data could also be done for other countries and rusts. The hope is that the technology will be a valuable weapon against the constant evolution of new disease strains.
The Sen4Rust demonstration is the first successful integration of landscape level wheat vulnerability maps with advanced epidemiological models.
It represents a significant step change in the fight against crop disease, and could advance disease forecasting — as well as our knowledge of the underlying pathogens.
Following EWAS expansion, the integration of Earth observation data could also be done for other countries and rusts. The hope is that the technology will be a valuable weapon against the constant evolution of new disease strains.
