blog

In recognition of World Food Day this post highlights two articles recently published in the OKAD channel that illustrate the power of using data-rich visualization approaches to meet the Sustainable Development Goals of eliminating hunger, malnutrition and poverty by 2030.

The DCL Atlas by Hyman et al.

Although climate change will negatively affect most of the world, some regions will be harder hit than others. This is especially true for the world’s dryland areas that border the tropics; where ever more extreme temperatures and droughts could have a devastating effect on the crops on which these communities depend. This is a truly massive problem as half of the world’s population, 60% of the world’s poor, and 60% of the world’s malnourished children live in dryland areas.

Dryland communities typically rely on hardier crops than global staples such as wheat, rice and corn; the main produce cultivated in these areas are the Dryland Cereals & Legumes (DCL): chickpea, common bean, cowpea, faba bean, groundnut, lentil, pigeon pea, soybean, barley, pearl millet, small millet and sorghum. Although they are better adapted to high temperatures and low precipitation than other cereals, it is unknown how well current varieties will tolerate the extreme weather variations that will become increasingly common.

Improving current and future food security in these areas is a big reason why the Dryland Cereals and Legumes Agri-Food System research program was launched by the CGIAR, the consortium of the world’s main agricultural research centres that played a major role in kick-starting the Green Revolution. But with finite resources and facing such a vast problem, they needed to identify the regions that would benefit most from their research investment.

The program commissioned geospatial researchers based at several CGIAR centres to produce evidence to help them focus their efforts. The group acquired and integrated a host of relevant data sources; from rainfall simulations to local earnings and nutrition indicators. These were then converted into pixels and overlain on a global map.

What they found was very helpful. Although they stress their findings should not be used to exclude certain regions from investment, they identified that the Sahel, East Africa, and South Asia would benefit most from DCL research considering their large populations, extent of land devoted to DCL crops, and high levels of poverty and malnutrition. Barley, pearl millet, sorghum, soybean and cowpea are the most common crops in these regions, so research into increasing their yield, testing their resilience to climate stresses, and biofortification with essential nutrients should generate the greatest benefits.

They also located regions where many different DCL crops are grown together, with the Sahel and the (previously) fertile crescent in the Middle East standing out in particular. This is useful information as location-dependent fixed research costs (infrastructure, salaries etc.) can be shared between initiatives focussing on different crops, which could lead to significant savings that could be invested elsewhere.

The team have made the database and geospatial visualization tools openly accessible at the DCL Atlas.

The CELL5M database by Koo et al.

Market accessibility based on travel time to cities with populations greater than 20,000.

CELL5M, another openly available geospatial database, restricts its geographical focus to Sub-Saharan Africa, but integrates and harmonizes a dizzying array of more than 750 data layers, plotted at a fine-scale resolution of 5 arc-minutes (each pixel roughly equals 100km²). The potential uses of the database, which was developed by IFPRI (a CGIAR centre) and the University of Minnesota, are practically limitless. It integrates almost every relevant data type imaginable: crop yields, market value, climate, pests, soil type, income, populations, health, you name it!

Although the paper describing the database is new and not peer reviewed at the time of writing, the database itself was actually compiled in 2010. It was originally built to power the MAPPR and TABLR visualization tools developed by HarvestChoice, a major mapping initiative that helps inform strategic decisions aimed to help improve the livelihood of smallholders in Sub-Saharan Africa.

Other researchers have made very good use of it as well. Since its creation, CELL5M has been instrumental in hundreds of projects ranging from mapping plant disease threats, estimating the time it takes farmers to travel to market (see picture) and the discovery that child growth stunting due to malnutrition is most prevalent in dryland areas.

The need for open agricultural data

Both articles illustrate the power of integrating and visualizing open datasets in the global struggle to improve the 750 million or so lives that make up the world’s poor. Unfortunately, many potentially useful datasets remain inaccessible or unusable, which underscores the great value of global initiatives, such as IGAD, GODAN and their partners, that advocate open agriculture and nutrition data.