ONE ACRE FUND MEL KE RW TZ
In-situ data were collected for Monitoring, Evaluation and Learning (MEL) via agronomic surveys in Kenya, Rwanda and Tanzania, 2016-2019. Data includes maize, beans, climbing beans and potatoes. On the quality the owner wrote the following. The scope of data collection differed by country and by year based on internal needs. Survey questions were not always uniform across countries. Many of these data rely on self reports by farmers, which is inherently imprecise. However, harvest estimates are based on physically weighed harvests from 2 randomly placed boxes, and can be considered fairly reliable.
Provider's website
Documentation
One Acre Fund (2020). Data from MEL agronomic surveys in Kenya, Rwanda and Tanzania, 2016-2019
Pre-processing
Original data filtered:
- excluded records with intercropping (intercrop = 0)
- excluded records with incomplete location
- excluded records with incorrect unique ID string
- excluded records with unrealistic short crop cycle length (<60 days)
- excluded records with unrealistic long crop cycle length (>300 days)
We estimated the spatial accuracy around 100 m based on the remark "registred latitude/longitude, though occassionally this will not have worked as intended and was taken from homestead or near the field".
CROP_DEV_BBCH = 00 (planting)
Attribute: planting date
Remark: farmer reported planting date
CROP_DEV_BBCH = 99 (harvest)
Attribute: harvesting_date
Remark: date harvest crop cut was conducted
SO_FWT_KGHA
Attribute: yield_kg_ph
Remark: assumed yield is expressed in fresh weight. Yield measure in the field using 6x6 m harvest boxes (placed diagonally) or 8x5 m harvest boxes (placed perpendicular to rows) though other options were possible if the plot was small or narrow)