Zonal Aggregation
Definition
Zonal aggregation summarizes raster or point data within predefined zones—administrative areas, watersheds, market territories—producing statistics like mean, sum, min/max, or percentiles. It translates continuous data into decision-ready indicators while retaining the geometry of interest.
Application
Public health aggregates exposure, agriculture summarizes rainfall by farms, utilities compute outages by district, and retailers roll up demand by trade areas.
FAQ
How do you avoid edge bias in coarse rasters?
Use area-weighted sampling or super-sampling at zone boundaries rather than simple cell inclusion tests.
When should medians or quantiles be preferred to means?
For skewed distributions (income, pollution spikes) where means are misleading; quantiles capture typical conditions and extremes.
How to handle small zones with few samples?
Report uncertainty or suppress unreliable estimates; consider hierarchical models that borrow strength from neighbors.
What lineage should accompany aggregated results?
Document source rasters, time windows, resampling choices, and zone definitions so others can reproduce the numbers.