1 of 8

Project Presentation

​

Team:

  1. Russo Mnene
  2. Edward Ouko
  3. Steve Firsake
  4. Clarine Kigoli
  5. Kennedy Were
  6. Leonard Sweta
  7. John Jaoko
  8. Marion Warau
  9. Stephen Korir

2 of 8

Problem Statement

Dynamic World Mapathon, Nairobi

Crop phenology : Establish the relationship between phenology response based on

sentinel 2 harmonized and dynamic world datasets probability prediction.

Study Area: Ahero, Chemelil, Moiben, Narok, Kericho

Classes of Interest :

1.Rice

2: Sugarcane

3: Maize

4: Wheat

5: Tea

3 of 8

Methodology

Dynamic World Mapathon, Nairobi

  1. Chose area of interest
  2. Selected 5 farm regions with known existing crop types for comparison
  3. Generated time series analysis for NDVI and DW crop probability values for both wet and dry seasons
  4. Compared phenology and how crops appear at different stages
  5. Looked at at the DW series to see how the classifier picked out crops in the varying stages
  6. Further compared dominant DW classes on the map with NDVI for the same seasons over the different farms.

​

4 of 8

Overall result

5 of 8

Results- Phenology for wet season

Dynamic World Mapathon, Nairobi

6 of 8

Dynamic world prediction

7 of 8

Results- Phenology for dryseason

8 of 8

Code

Split panel code:

https://code.earthengine.google.com/0012e1e4264177b0bf852c42662eed92

​

App code:

https://code.earthengine.google.com/116a525dc2e36dffc0ef75dc49a2caa8

​

​

​

​