We are looking for a master student who will develop an approach to identify and count trees on cocoa farms in Ghana based on satellite images. The project involves 2 field trips to Ghana and a very hands-on experience of the use of satellite images to calibrate and validate a model that can be applied in the cocoa industry. In the first trip, the student will take a sample of farms to calibrate a tree-counting model; and during the 2nd trip he/she will validate whether the model works with another sample of farms. Flight to Ghana, within-country travel expenses and accommodation costs are covered by the university.
Cocoa farmers in Ghana cannot get access to finance to improve their farming practices. Financial institutions want to find a validated and cost-efficient manner to estimate cocoa farmers’ income to determine whether to extend loans to farmers. Cocoa farmers’ income is based on their productivity, which is based on the number of trees a farmer has on his/her farm (among other factors).
The master student should:
- Select a sample of 10 farms, where you will collect a great amount of detail: count the number of trees, tree height, size of canopy, discuss with farmers and make notes on what is the age of these trees, on whether the farmers prune these trees, use fertilizer, which seeds they use (for young trees) etc. You will define the exact questions for farmers together with Diana Kos (the full baseline survey is complete – we just want farm-specific questions).
In the Netherlands:
- Compare your baseline with satellite images of those same sample farms, compare that to your farm notes from the field, and based on that, develop a model to:
- estimate the number of trees in farms, and
- Identify cocoa vs non-cocoa trees on farm. The size of the tree canopy and visibility of trees will be influenced by farmers’ pruning, use of fertilizer, seed type he used on farm etc.
- Use the results to validate the model with the second field trip.
- Use this calibrated and validated model for whole region estimation.
- Strong analytical skills
- ArcGis knowledge
- Able to go to fieldwork
- Communication skills
Theme: Integrated Land Monitoring