GIS Farming in Africa

Claire Jacobs¹ and Kees van ‘t Klooster², Alterra, Lusaka¹ Wageningen², the Netherlands

FARMD (March 2012) | Satellites are used for monitoring and recording data on a wide variety of parameters such as rainfall, temperatures, crop variables, and soil and water parameters. This information can be easily obtained on a large scale. This is especially useful for Africa, where little infrastructure is available to collect all the information by land-based methods. When using remote sensing, calibration and validation of the received signals with ground data is essential for reliable and realistic correlation between the Remote Sensing (RS) signal and the actual parameter of interest. It might be appealing to start using the RS data with little or no calibration and validation, especially in the context of Africa, where reliable ground data may be hard to find. However with responsible handling of the data, this information can leapfrog agricultural development in Africa and is well worth full exploration and exploitation. These information techniques can be used in two ways: As planning tools or as operational tools.

Planning tools

When used for planning, the information is processed to produce outputs which can assist agricultural actors in their decision making and planning processes e.g. crop suitability maps, yield maps, harvest prediction maps, weather risk maps etc. GIS and Remote Sensing are combined with domain knowledge, amongst others soil, climate, hydrology and irrigation. Innovative expertise is available in the building of, and advising on, spatial data portals and clearing houses, to convert plain data into project needed knowledge. A typical example is yield forecasting on country level in various continents, including Africa. If countries know well in advance the size of their yields they can plan accordingly.  For example,  a country forecasting a shortfall in a specific crop yield might in advance order additional supplies in from abroad, thereby reducing the adverse impact of a shortfall in production.  Alternatively a country anticipating a bumper harvest may pre-emptively start arranging additional export contracts or preparing additional storage facilities.  Countries may also utilize such advance information to assist them in their pricing schemes and programs enabling them to tailor program details by better understanding likely production that season.

Weather risk maps

Another example of the use of remote sensing techniques are weather risk maps. In rain-fed agriculture, the income of farmers can be at risk of severe floods and droughts. However the chances of floods or droughts occurring in an area can be calculated. High yield risks cause farmers to be extremely cautious when investing in inputs that may be wasted. This is one of the reasons for low yields and low increases in yields for many subsistence farmers in Africa. Reducing such risks by spreading/pooling such risks over a large group of farmers through insurance schemes is an interesting option for reducing the individual risk of a farmer. With more income stability the farmer can better afford to invest in inputs and raise his average yields, as the risk of is substantially reduced. However designing appropriate risk distribution schemes is far from easy. Yields in agriculture, be they high or low, are usually caused by a combination of factors. In some rain-fed agricultural systems in Eastern Africa a significant correlation between rainfall and yield does exist e.g. a strong correlation between rainfall amounts / frequency and maize yields can be found in a number of East African locations. In such cases an insurance scheme based on rainfall is practical and relatively easy to verify. However in other cases the correlation between rainfall and yield is much less clear, and the  regression does not capture the relationship, as the relationship between weather and yield is not a straightforward one. A combination of poor management practices, soil organic matter, lack of credit or other inputs is causing strong losses in yields. In such cases a more integrated approach provides farmers with a tailored support package, where the provider of the support package is also sharing the risks and the payment by the farmer for the package depends on the final yields. This delivers incentives for both the farmer and the provider of the services to maximize the yields.  

Operational tools

Precision farming is a concept where site specific information is used to assist operations in the field. All the features of a specific location including soil characteristics, water availability, past and present crops grown at that location and past, present and forecasted weather conditions are held within the system as all of these features have an influence on yields. This data is then analysed by the system to provide useful information to agricultural actors on the ground regarding the specific location and its suitability for growing certain crops.  Essentially if such data are stored in a Geographical Information System (GIS), the user has the basic information that can be processed by an expert system to determine the most optimum next step. For example, if you know that at a certain spot the soil is very fertile and extra high yields are possible, provided that the plant nutrients are in sufficient amounts available, you may want to dose some extra fertilizer, in order to realize the full potential. On the other hand in a place where no seeds have germinated, why would you waste money by putting fertilizer on such a spot?  Such a tool enables farmers and agribusinesses investors to improve their understanding of what crops they can produce most efficiently in those locations, and how they can minimize risk and maximize returns.

African traditional farmers know their fields well, and act in site-specific ways accordingly. In many places traditional subsistence farming methods are being replaced by more mechanized systems, in order to boost production.  Such machinery may be useful, but also tends to treat all land in all areas equally. This can result in under-watering in some places and over-watering in other places – as key site-specific information is not considered. However by utilizing GIS expert systems it is possible to build ‘brains’ into such machines, enabling adjustable dosing for water, fertilizers and pesticides, allowing a producer to maximize efficiency i.e. by optimizing the use of inputs to achieve maximum yields at low costs. An added benefit of such an approprach is that excess use of fertilizer or pesticides is avoided and environmental risks and water pollution can be minimized.

Where the costs of the use of these precision farming techniques can be covered by reductions in input costs and increases in yields, the introduction of this technology has a good future. The enabling environment to apply such technology is already partly there with the current satellite systems, and that part will grow in the years to come. A bigger challenge for use of this technology is the more down to earth side of it: To have the infrastructure to finance the machinery and the maintenance to keep things up and running, as well as the knowledge systems needed to interprete and act upon the collected data is the more difficult part of the equation. Early examples have shown that precision farming can be effectively done in Africa. The challenge moving forward will be to work to ensure that precision farming can ultimately enter mainstream agriculture in Africa.