National food security monitoring

Sovereign crop and rangeland monitoring gives governments production estimates before markets move, replacing commercial data subscriptions that can be withdrawn or priced out of reach.

The dependence this ends: Import decisions made on last season's guesswork

The information gap that precedes every food crisis

Most governments learn that a harvest has failed from the same source as traders: a commodity index published in Chicago or London, weeks after the growing season closes. By then, import prices have moved, neighbours are competing for the same shipments, and the minister is managing a crisis rather than preventing one. The asymmetry is not accidental. Commercial crop-intelligence products are sold to the buyers who pay most, and those buyers are not governments of food-importing nations.

The post-2022 environment has made the dependence sharper. Commercial satellite operators have demonstrated, in Ukraine and elsewhere, that imagery access can be restricted over contested areas. Index providers use composite data whose provenance a sovereign buyer cannot audit. A government that cannot independently observe its own growing regions is, in practice, outsourcing its food-security intelligence to parties with different interests.

What the satellite actually measures, and what it cannot

Multispectral imagers in sun-synchronous orbit measure reflected light across visible and near-infrared bands. From those measurements, analysts derive vegetation indices, principally NDVI and EVI, that correlate reliably with crop biomass and stress. When combined with GNSS radio-occultation data, which retrieves atmospheric temperature and moisture profiles without active radar, the system can detect drought onset at regional scale before it is visible on the ground. That combination underpins the crop-estimation work Satellize delivered for the Kingdom of Tonga.

Honest limits matter here. Optical sensors cannot see through cloud. In humid tropical zones with persistent cloud cover, revisit frequency must be high enough to guarantee cloud-free composites within a ten-to-fourteen-day window, which typically requires multiple satellites or a planned data-fusion arrangement with open-access sources such as ESA's Sentinel-2. Spatial resolution for field-level mapping requires ground sampling distances below ten metres; coarser imagery is adequate for regional production estimates but will miss smallholder plot variation. The system tells you crop condition and probable yield range. It does not tell you post-harvest losses, storage capacity or distribution logistics.

The ambition ladder: from pathfinder to operational programme

A pathfinder mission is a single multispectral satellite, typically in a 500-to-600-kilometre sun-synchronous orbit, paired with a ground station and in-country processing capability. It proves the data pipeline, trains the national team, and produces one to two full growing-season analyses. Small-satellite missions of this class have publicly reported budgets in the low tens of millions of dollars, with programmes such as the UK's NovaSAR-1 and various national Earth-observation pathfinders offering a rough public benchmark. Timeline from contract to first operational data is typically twenty-four to thirty-six months, depending on whether a heritage bus platform is used and how quickly spectrum coordination proceeds through the ITU.

An operational constellation adds two to four satellites to reduce revisit from several days to under forty-eight hours for any given field location. At that revisit cadence, the system can track crop development through phenological stages rather than just snapshot the end-of-season state. It can also support drought early warning with enough temporal resolution to trigger import-procurement decisions six to eight weeks before a shortfall becomes visible in market prices. Programmes at this scale, such as India's Resourcesat series or the European Copernicus contributing missions, involve multi-year procurement cycles and correspondingly larger budgets; the public record does not support a single tidy number, but constellation programmes of four to six small satellites with ground infrastructure have been publicly reported in the range of tens to low hundreds of millions of dollars across a five-to-seven-year programme arc.

The distinction between the two levels is not just resolution or revisit. It is institutional. A pathfinder can be operated by a small technical team with external analytical support. An operational constellation requires a national geospatial intelligence function: people who can interpret outputs, brief ministers and defend methodology when the numbers are politically inconvenient. That institutional investment is at least as important as the hardware.

What the customer owns at handover

Sovereignty in this mission means owning the observation, not licensing it. At handover, the government holds the satellite itself, the ground station and its software stack, source-access rights to all processing algorithms, and the trained national team that runs daily operations. Crop-model outputs, historical archives and validation datasets are held in-country on infrastructure the government controls.

What remains with Satellize or its integration partners after handover is limited to warranty obligations on hardware and, if the customer elects it, an optional technical-support retainer. The analytics methodology is documented and transferred, not kept as a proprietary black box. That matters when a new government takes office and wants to audit how last year's production estimate was derived.

The analytical layer is not optional

Raw satellite data is not a food-security assessment. The gap between a calibrated reflectance image and a number a minister can use in an import-tender decision is filled by exploitation software, validated crop models, and analysts who understand both remote sensing and local agronomy. Satellize's Tonga crop-estimation programme demonstrated that this layer can be built and transferred to national teams; it also demonstrated that the transfer requires deliberate curriculum design, not just software installation.

In-country processing is a sovereignty requirement as much as a latency one. If imagery must leave the country to be processed, the analytical independence the programme is meant to create is partially surrendered. The architecture should route raw data to national infrastructure first, with external calibration support available but not mandatory for daily operations.

What this mission is built from

Multispectral imagers: Primary sensor for vegetation index derivation, crop-stress detection and seasonal production estimation.
GNSS radio-occultation payloads: Atmospheric moisture and temperature profiling to support drought early warning and improve crop-model inputs.
Sun-synchronous orbits: Consistent solar illumination angle across repeat passes, essential for reliable multi-temporal vegetation index comparison.
In-country data processing: National data infrastructure that keeps raw imagery and derived assessments under sovereign control from reception onwards.
Exploitation and analysis software: Crop-model execution, yield-range estimation, drought-indicator dashboards and the analyst tools that translate data into ministerial briefs.

What you end up owning

The satellite or satellites, with full hardware audit rights exercised before and at handover
The national ground station, including antenna, command-and-control software and data archive
Source-access rights to all processing algorithms and crop models, documented for independent audit
Historical imagery archive and all derived analytical products from the programme's inception
Trained national operator and analyst team capable of independent daily operations
Exploitation software licences on terms agreed before contract signature, with no ongoing per-seat fees payable to Satellize

Handover is staged across the programme arc: the national team operates alongside Satellize engineers for at least one full growing season before sole-responsibility transfer. At handover, all credentials, documentation and source code are transferred under terms agreed at contract signature, not negotiated at the point of dependency. Satellize retains no ongoing data access and no licence over outputs; optional post-handover support is a separate, time-limited retainer the customer may or may not elect.

Programme parameters

Pathfinder configuration	1 multispectral satellite, 1 ground station, in-country processing node
Operational constellation	3 to 5 multispectral satellites, 1 to 2 ground stations, expanded processing infrastructure
Orbital regime	Sun-synchronous, 480 to 620 km altitude typical
Ground sampling distance	5 to 30 m depending on imager selection; field-level mapping requires ≤10 m
Revisit (pathfinder)	3 to 7 days per location; cloud-free composites within 10 to 14 days in most climates
Revisit (constellation)	Under 48 hours per location, enabling phenological-stage tracking
Indicative programme timeline	Pathfinder: 24 to 36 months to first operational data; constellation: 48 to 72 months to full operational capability
National team to operate	Pathfinder: 4 to 8 trained operators and analysts; constellation: 12 to 20
ITU coordination	Required; frequency filing should begin at programme inception; coordination adds 12 to 24 months if not started early
Cloud-cover limitation	Optical sensors cannot penetrate persistent cloud; multi-source data fusion with open Sentinel-2 archive recommended for high-humidity zones

Sources

One contract, one accountable engineer

Commissioned as one programme, not a stack of contracts: spacecraft, launch, ground segment, mission control, training and handover are priced together. Source-access terms and audit rights are agreed in writing before signature. Request a pathfinder scoping review.