Skip to main content
Regional Forecast rolls every field in an administrative region (state, division, district) into one weekly outlook. It is the layer between Estate Group and National Overview, and it is the primary view for agencies, cooperatives, and regional planners who do not operate individual fields but do commit resources across many of them. Every number on this dashboard is defined by the Aggregation Model. Confidence chips, freshness badges, and drilldown rules all follow that contract.

​
What the dashboard shows

Regional coverage

Total hectares under monitoring in the region, coverage as a percentage of official planted area (from agency records), and count of estates contributing data.

Production forecast

Central estimate (tonnes) with P10 to P90 confidence band. Broken down by crop and by the 14-day, 30-day, and end-of-season horizons.

Health distribution map

Choropleth of the region colored by area-weighted health score. Cells with fewer than 5 fields collapse into “Other” for k-anonymity.

Top hazards

The five rule cards firing across the largest area in the region this week, with total hectares affected and week-over-week trend.

Weather stress outlook

14-day weather-driven risk projection: heat units, rainfall deficit or surplus, and flood or drought watch zones.

Response rate

Share of alerts issued in the region that received a scout task, VRA action, or verified follow-up within their mitigation window.

​
How the forecast is produced

The central estimate is an area-weighted sum of per-field yield projections. The confidence chip is the minimum of every input’s confidence, so the weakest coverage or freshness signal drives the overall label. See Aggregation Model - Confidence degradation for the exact rules.

​
Refresh cadence

All cards show Last updated and Data through. A stale card is always shown, never hidden.

​
Access and privacy

Regional dashboards enforce k-anonymity with k = 5 on every drilldown. If a slice (e.g., a district for a niche crop) has fewer than 5 fields, that slice collapses into “Other” until the threshold is met. Aggregate totals still include those fields; only per-slice drilldown is suppressed. Regional users never see per-field identity for fields outside their own estate group. This is enforced at query time, not just at display time. See Aggregation Model - Privacy and access rules.

​
Common workflows

Open the region on Monday morning. Read the top-hazard card and the 14-day weather outlook. If any hazard’s affected area is growing week-over-week, coordinate a bulk scouting sweep with member estates. Export the dashboard as PDF for the member meeting.
When the weather stress card flags a flood watch zone, click through to the affected fields. The regional export bundles all field-level Verification records for the impacted area, ready for pre-emptive NADMA submission.
Filter production forecast by variety group. The regional total feeds subsidy allocation decisions. Because every input field is verifiable via Verification Model, the number holds up in the funding committee.
Compare response rate across estates in the region. A persistently low response rate on High alerts is a signal for extension officers to intervene. Drilldown respects estate-level privacy: you see the estate name and rate, not the specific fields.

​
What is deliberately excluded

  • Field-level rule cards - agencies see rule card counts and affected area, not individual field diagnoses.
  • Named scout assignees - regional users see task counts and completion times, not who did the scouting.
  • VRA prescriptions - regional users see “X ha with VRA applied this week”, not the maps themselves.
If an agency needs field-level detail (e.g., for a specific NADMA claim), it flows through Verification bundles issued by the estate, not through this dashboard.