It was a timely operation that influenced our best practices while dissecting challenges caused by climate change. Now we have an overview of the inventory, helping us to find ways to achieve our anticipated yields. The results have supported our decision-making and response to the terrain challenges on our farm.
Uganda Flying Labs helps mitigate risk for coffee farmers
Coffee is Uganda's largest export crop, ranking seventh in global coffee production and second in Africa behind Ethiopia. Uganda contributes three percent of the global output of both the arabica and robusta varieties. Arabica coffee beans tend to have a smoother, sweeter taste, while robusta beans have a more bitter taste. Typically, countries harvest coffee beans once a year, but in Uganda, the harvest period is from November to February, and the fly crop—the smaller, second harvest—from June to September.
Due to climate change, coffee crops are highly vulnerable. Rising temperatures and increasingly erratic rainfall are already exposing trees to more pests and diseases and decreasing both the volume and quality of crops. Furthermore, the COVID-19 pandemic affected the yield because most farmers faced challenges in cash flow to support their activities for good yield outcomes. In addition, commercial and rural farmers often lack access to essentials like fertilizer, irrigation equipment, and high-quality seeds; finance, agricultural training, and market data; and infrastructure like paved roads and processing facilities.
The Uganda Coffee Development Authority (UCDA) is a government agency responsible for maximizing coffee yield production, export, and quality control in the country and lenders. To mitigate the risks of poor crop yields and ensure repayment of agriculture loans UCDA must conduct research to monitor world market price changes and communicate with other international organizations to promote Uganda’s coffee to the world. Loans are available to finance soil conservation, irrigation, cultivation, machinery, and more. To help farmers, there is a need for adequate real-time data to address appropriate decision-making or rapid response to crop loss, which ultimately affects the Ugandan economy.
Conventionally, monitoring of crops is carried out by on-the-ground physical inspection by the farmers or experts (agricultural support staff deployed by the government). However, this is often expensive and time-consuming, and the assessments are done only in sample areas, which may not be fully representative of the situation in the entire region. Also, inspections may happen when the damage has already been done, for example, when there is an outbreak of a disease or pest infestation.
Depending on their overall health, plants reflect different amounts of visible green and near-infrared (NIR) light. These variations can be captured from imagery and analyzed.
Through funding from a microgrant for turning data into action, Uganda Flying Labs, a company that specializes in drone, mapping, and data analytics and aims to promote sustainable adoption of technology in the region embarked on a project to capture multispectral and RGB (red, green, and blue) imagery via drones with cameras to help spot crop health issues that are often invisible to the human eye alone.
The scope of the project included a deep learning workflow using Esri's ArcGIS Image Analyst and ArcGIS Pro tools to map the individual stands of coffee plants, perform a health assessment, and indicate areas under stress. ArcGIS Image Analyst provides a variety of tools for advanced image interpretation and geospatial analysis. The project focused on monitoring the robusta coffee species, which is suitable for low-altitude areas and tropical climates within the Lake Victoria basin.
Using a DJI Phantom 4 Multispectral drone (with an RGB camera and five monochrome cameras), 44 hectares of coffee plantation were captured in several hours. From this imagery, a digital surface model, reflectance maps for the various bands, and a 3D point cloud were produced. Using ArcGIS Pro, a digital terrain model slope map, contour maps, and normalized difference vegetation index (NDVI) maps (an indicator of crop health) were generated.
The RCNN Mask deep learning model, available in ArcGIS Pro, was trained and run to detect and map 7,000 coffee plant stands.
Each crop's status was then assessed by overlaying the NDVI map and identifying areas with stress.
The slope and contour maps aided in assessing erosion and flood risks.
The results of the project were well received by all stakeholders.
UCDA, as a national body concerned with reporting figures pertaining to crop yield, was particularly interested in seeing how this technology can help in estimating yields nationwide and efficiently address factors limiting improvements in those yields and, subsequently, exports.
Knowing the total number of crop stands, harvest estimates were calculated based on a mature coffee tree's expected yield per year. On the approx 7.5ha site that the drone was flown over, the harvest was estimated at about 8,500 Kg of green coffee beans (dried and without husks). .
Understanding the risk of erosion in areas with high-value crops like coffee will also help UCDA work with farmers on implementing techniques for retaining soil on slopes and restoring nutrients.
For lenders and insurers, understanding the expected volume of yield, the erosion and flood risk, and the number of trees that need replacement due to disease helps with protecting their investment, assessing return on investment, and identifying where investment is needed via loans given to farmers.
Using the crop health maps, farmers had an overview of the overall health of their stands and could perform harvest projections and see if any trees needed replacing. In addition, contour maps contributed to the farmers' thinking of how to construct drip lines.
The result of the project was a win for farmers, exporters, and lenders. They can use an intuitive, easy-to-understand GIS application to view and analyze the data, and generate actionable information in relation to crop stress, soil health, water needs, erosion, and flood risk so that the farmers can quickly intervene when needed.
Agriculture in Africa
Learn more about the Esri Geospatial Program for Agriculture in Africa and modernizing agricultural workflows using GIS. Go to agriculture.africageoportal.com.