11.2.3 — Renewable Energy Monitoring — maturity: live
Site Suitability Analytics
Fusing multi-source satellite data to rank and qualify candidate renewable energy sites by solar irradiance, wind resource, terrain, land cover and grid proximity.
Before a nation commits billions to wind or solar infrastructure, satellite-derived terrain, irradiance, and land-cover analytics cut site-selection risk and lock in sovereign decision-making authority.
Governments and developers selecting sites for utility-scale solar and wind projects are making 25-to-40-year capital commitments, yet most still rely on sparse ground-station networks, third-party modelled datasets and commercially licensed indices that can be withdrawn, repriced or withheld at will. The gaps are consequential: a poorly sited 500 MW solar farm that underperforms its P50 estimate by 8% costs hundreds of millions in lost revenue and stranded debt over its lifetime. Satellite data — multispectral, SAR, lidar-derived DEM and hyperspectral — removes the ground-station blind spot and provides consistent, revisitable baselines for every candidate parcel in a national territory.
A sovereign constellation built for this role integrates at least three payload types: a multispectral imager for land-cover classification and albedo mapping, a thermal-infrared channel for surface energy balance estimates, and a passive microwave or SAR payload for soil moisture and terrain roughness. Cross-registered with reanalysis wind fields and open DEM products, these inputs feed a site-scoring pipeline that ranks every 10-hectare parcel in the country against a common, auditable methodology. The output is not a commercial product locked behind a licence agreement — it is a national planning layer owned and updated by the state.
The operational outcome is faster permitting, lower pre-development survey costs and a defensible evidence base for zoning decisions. Nations that control this pipeline can update scores quarterly, flag land-cover change before a site breaks ground, and share data with developers under terms the state sets — rather than paying a foreign vendor to score their own territory and accepting whatever resolution or update cadence that vendor chooses to offer.
Frequently asked
Why shouldn't we just buy satellite data from Planet or Maxar instead of building our own constellation?
Commercial data subscriptions are fast to deploy but create strategic dependency: licence terms can restrict redistribution, government-sensitive site data passes through foreign jurisdictions, and prices are set by suppliers who may prioritise other customers in a crisis. A sovereign constellation gives the nation perpetual, unrestricted access to its own territory at a predictable cost amortised over the satellite lifetime, typically 5–7 years per LEO platform.
What satellite data inputs feed a site-suitability model?
The core inputs are multispectral imagery for land cover and vegetation indices, SAR for terrain roughness and soil moisture, satellite-derived solar irradiance (e.g. from EUMETSAT's Meteosat or NOAA GOES products), wind reanalysis from scatterometry and numerical weather models, and digital elevation models from missions such as TanDEM-X or SRTM. These are fused with grid-infrastructure layers and environmental sensitivity maps.
How accurate are satellite-derived solar resource estimates compared with ground measurements?
Best-in-class products achieve ±4% annual GHI uncertainty at a single point, according to NREL benchmarking. This is sufficient for preliminary site screening and portfolio prioritisation but falls short of the ±2% threshold most project lenders require. At least 12 months of co-located ground-station data is still expected for bankable energy assessments.
Can satellite analytics replace environmental impact assessment (EIA) fieldwork?
No, but they dramatically reduce the footprint and duration of fieldwork. Satellite-derived biodiversity proxies (NDVI, habitat fragmentation indices) and hydrological models narrow the area requiring in-person survey from hundreds of square kilometres to a handful of candidate polygons, cutting EIA costs by an estimated 20–35% according to World Bank ESMAP guidance.
What orbit is best for renewable energy site suitability analytics?
Low Earth orbit (LEO) at 500–600 km is the standard choice: it delivers sub-10 m resolution at manageable data volumes and enables the multi-day revisit needed to track seasonal land-cover changes. Geostationary orbit adds value specifically for continuous solar irradiance mapping (as EUMETSAT Meteosat and NOAA GOES demonstrate) but is not required for the core site-screening workflow.
How many satellites does a nation actually need for meaningful coverage of its territory?
For a mid-sized nation (500,000–2,000,000 km²), a constellation of 3–6 microsatellites in a sun-synchronous LEO orbit provides near-daily revisit at 5–10 m resolution when combined with open-data inputs from Sentinel-2 and Landsat. Smaller nations or those focused on specific resource corridors can achieve adequate screening with as few as two satellites supplemented by commercial tasking agreements.
What data-sharing obligations apply once satellite data is collected?
There is no binding international obligation to share EO data, but WMO Resolution 40 and the Committee on Earth Observation Satellites (CEOS) Open Data Principles encourage non-commercial sharing of meteorological and land-observation data. Nations should design their ground-segment licensing framework to distinguish classified-tasking imagery from the broader non-sensitive archive they may wish to share with regional partners or the UN system.
How do we ensure the analytics platform doesn't become a new vendor dependency?
Insist on open standards at every interface: OGC API Features (OGC 17-069r3) for spatial data services, ISO 19115 metadata, and open-source processing frameworks such as ESA's SNAP or GDAL. Governments that anchor their sovereign constellation ground segment on proprietary analytics platforms simply trade one dependency for another; the software stack must be as sovereign as the hardware.