Ethiopia’s main economic activity is Agriculture. Smallholders in Ethiopia farm around 37% of the lands and the rest 8% goes to the large-scale commercial farms. Large-scale commercial farms produce crops such as rice, maize, coffee, tea, cotton, pulse, rubber, and palm oil.

Invest in agriculture is the most important and most effective strategy for poverty reduction in rural areas. As part of Ethiopia’s development strategy, the Ethiopia government has set out to attract more foreign investment in large-scale commercial agriculture as outlined in this policy.

Mapping Commercial farms with the World Bank

In the large commercial farm mapping task with the World Bank, we mapped more than 190 large commercial farms in lowlands Ethiopia. Ethiopia lowlands, usually defined as places below 1,500 meters in altitude, account for approximately 60 percent of Ethiopian territory and 12 percent of the population.

Mapping process:

Big farmlands can be visible from high-resolution satellite imagery. But such satellite imagery in OpenStreetMap for Ethiopia is outdated and high cloud cover. These factors make mapping difficult.

To better locate large commercial farmlands, we used multiple satellite imagery and tools: High-resolution Satellite imagery (e.g. ESRI base map), Sentinel-2 Agriculture mosaic, and Java OpenStreetMap editor (JOSM).

• Satellite Imagery ESRI World Imagery was used for mapping because it is up to date than others in Ethiopia lowlands

• Sentinel-2 Sentinel-2 is open source imagery provided by [ESA] (https://en.wikipedia.org/wiki/European_Space_Agency), it’s in 10m resolution with multi- spectral bands. It used to perform terrestrial observations like forest monitoring or land cover changes detection. The imagery is updated every 16 days. To make agriculture band mosaic, band 11, 8 and 2 combination specifically to spot farmlands.

• Tool for mapping JOSM is a desktop application that allows mappers to trace, map and manipulate data easily, thanks to its plugins and stability. For mapping were used ESRI and Sentinel-2 images both as base layers in JOSM. After 2 weeks of mapping, 194 commercial farms were added to OpenStreetMap. Before and after mapping: Previously, 55 farms were mapped in OpenStreetMap within Ethiopia lowlands. After, we mapped 194 farms in Ethiopia lowlands.

In the following table, you can see the number of farms added in Ethiopia’s lowlands according to size farm in hectares. We are looking for some validation, and Let me know if you have any questions and feedback!