Check the date and coverage first
Clouds, shadows, missing data and processing can affect the result. Grey areas without accepted data are not NDVI zero.
The same satellite does not guarantee identical results
The boundary, date, image selection, mask and area weighting must match for a meaningful comparison.
A map is not an automatic fertilizer rate
Fertilizer decisions also require the crop, growth stage, soil condition and expert review. A display grid does not create a more precise satellite measurement.
1. Field and season
The saved boundary defines the area used for calculations. Each image is linked to the field, season and boundary version so older data is not silently applied to a changed field.
2. Date, mask and NDVI
NDVI uses red and near-infrared bands. The mask separates accepted data from clouds, shadows and missing measurements. Missing coverage stays missing instead of becoming an invented zero.
3. Maps, cells and comparison
The source raster and field grid serve different purposes. Cell values are weighted by their intersection with accepted pixels. Date comparisons use common accepted coverage and show how much area is comparable.
4. History and additional data
Up to three years of archive searching is on the roadmap. Scenes and cloud conditions are checked by field and date, then processing runs in bounded batches. Free sources come first; an additional paid source requires a suitable need, a visible quote, consent and a spending limit.