Prerequisite 1 of 21

Agri zones by dominant crop

← All prerequisites
1Partially built

The 2015 proposal

Divide the country into agri zones by dominant crop, so advice, inputs and planning fit local soil, water and climate.

From the original 2015 deck

India’s agricultural regions — the basis the 2015 deck proposed for dividing the country into agri zones by dominant crop.
India’s agricultural regions — the basis the 2015 deck proposed for dividing the country into agri zones by dominant crop. · Source: ICAR

Where it stands in 2026

India has not one zonation but three overlapping ones: the Planning Commission-era 15 agro-climatic zones, ICAR-NARP’s 127 agro-climatic sub-zones, and NBSS&LUP’s 20 agro-ecological regions refined on soil, rainfall and growing-period data. What was missing for sixty years was resolution: a zone the size of a district cannot tell one farmer’s field from the next. The Digital Agriculture Mission is now closing that gap from below, building 1:10,000 soil-profile maps (about 142 million hectares planned, roughly 29 million mapped so far) and a geo-referenced crop-sown registry through the seasonal Digital Crop Survey. For the first time the macro zone and the individual parcel can be joined.

The open gap

The zones and soil maps exist; the per-parcel advisory layer that turns them into a sowing call for one farmer’s field does not — yet.

The path to close it

The zones and the soil maps already exist; the missing layer is turning a static map into living, per-parcel advice. Fuse the 1:10,000 soil grid and the 127 sub-zones into the Farmer Cockpit so every farmer gets a sowing-window, crop-suitability and variety call for their exact field — not a district average. The Microsoft–ICRISAT AI Sowing App showed the payoff is real and cheap: a single well-timed SMS lifted groundnut yields 10–30% with no new hardware in the farmer’s hands. Three tasks finish it — complete the soil mapping (about 29 of 142 million hectares done), refresh advisories each season against the live forecast rather than a long-run average, and expose the whole zone-and-soil layer as an open API any advisory app, public or private, can read. Then the zonation stops being an atlas and becomes an instruction.

Specifications — what “built” requires

Illustrative — a proposed specification and sequence, not an official government roadmap.

Acceptance criteria

  • Every operational holding is mapped to its agro-climatic sub-zone and its 1:10,000 soil-profile grid cell.
  • Each zone carries a current, season-specific sowing-window and crop-suitability advisory, refreshed against the live forecast.
  • Soil mapping is complete nationwide (the full ~142 m ha, not ~29).
  • Zone, soil and weather data are exposed as one open API any advisory app can read.
  • A farmer can pull a per-parcel sowing and variety recommendation for their exact field, in their language.

Technical spec

Zonation
ICAR-NARP 127 sub-zones × NBSS&LUP AER, joined to the 1:10,000 soil grid
Resolution
per-parcel (land-record survey number), not district average
Inputs
IMD/forecast feed + Soil Health Card readings + crop calendar
Output
sowing-window + crop/variety suitability score, per parcel, per season
API
open, versioned read endpoint (zone + soil + suitability)
Refresh
each season, against the live weather forecast

Roadmap to built — phase 1 → 2 → 3

Illustrative — a proposed specification and sequence, not an official government roadmap.

  1. 1
    Phase 1 · Now

    Complete the base layer

    Finish the 1:10,000 soil mapping and join it to the 127 NARP sub-zones.

    A complete per-parcel soil-and-zone base layer.

  2. 2
    Phase 2 · 6–12 months

    Make it dynamic

    Fuse that base layer with the live IMD forecast into a per-parcel sowing-window engine.

    Seasonal advisories replace the static atlas.

  3. 3
    Phase 3 · 12–24 months

    Open it up

    Expose zone + soil + suitability as an open API and deliver per-parcel calls in-language through the Cockpit.

    Any advisory app can serve any farmer — built.

See it working: Farmer cockpit →

Sources