Seeing the Isle of Man from Space
The Biosphere Observatory now includes three new Earth observation datasets that let you explore how the Isle of Man's land surface has changed over the past 25 years. Each one draws from different satellite missions and reanalysis products, and each tells a different story about our island.
Tree Cover: Hansen Global Forest Change
The first new dataset comes from the Hansen Global Forest Change project — a collaboration between the University of Maryland, Google, NASA, and the USGS. It uses 30-metre resolution Landsat satellite imagery to map tree cover globally.
We download the raw GeoTIFF raster tiles (each one around 200MB), extract the Isle of Man bounding box, and compute:
- Baseline tree cover (year 2000) at a 10% canopy threshold
- Annual tree cover loss from 2001 to 2024 — detected as stand-replacement disturbance
- Tree cover gain from 2000 to 2012
The page includes a reference comparison table showing how our satellite-derived figures compare with official sources: the World Bank/FAO (6.07% forest area), Global Forest Watch (1.7% at 30% canopy), the IoM Government's DEFA estate figures (~3,050 ha), and the Manx Wildlife Trust's State of Nature 2024 assessment.
Different definitions of "forest" produce very different numbers — and that's actually one of the most interesting things about this data.
Soil Moisture: ERA5-Land Reanalysis
The second dataset taps into ERA5-Land — the European Centre for Medium-Range Weather Forecasts' land-surface reanalysis at approximately 9km resolution. We access it via the Open-Meteo archive API.
For a central Isle of Man location, we pull daily soil moisture at three depths:
- 0–7 cm — the surface layer, most responsive to rainfall
- 7–28 cm — the root zone for grasses and shallow-rooted plants
- 28–100 cm — deeper soil, slower to respond but critical for agriculture and tree health
Alongside moisture, we also capture soil temperature, which reveals frost patterns and growing season length. The data runs from 2000 to the present day, aggregated into monthly, annual, and seasonal summaries.
Land Surface Temperature: MODIS + ERA5
The third dataset combines two complementary sources:
MODIS MOD11A2 from NASA's Terra satellite provides actual 8-day composite land surface temperature measurements at 1km resolution. These are direct radiometric measurements — the satellite literally measures thermal radiation from the ground. The catch? Cloud cover. The Isle of Man is famously cloudy, so many composites have gaps.
ERA5-Land fills those gaps with continuous modelled temperature data — air temperature at 2 metres, daily max/min, and soil temperature. It's not a direct satellite measurement but it's continuous and consistent.
Together, they tell a powerful story about warming trends, declining frost days, and changes to the growing season on the Isle of Man over the past quarter century.
Three New MCP Tools
As with every dataset on Smart Island, these come with MCP (Model Context Protocol) tools so AI assistants can query the data directly:
tree_cover— baseline, annual loss, gain, and net changesoil_moisture— multi-depth soil moisture and temperature, filterable by yearland_temperature— MODIS LST, ERA5 temps, frost days, growing degree days
All three are available via the Smart Island MCP server at mcp.smartisland.im.
Open Data Downloads
All three datasets are available as free JSON downloads from the Data Hub, alongside the existing 25+ datasets. Tree cover, soil moisture, and land temperature data can be downloaded in JSON, GZ, or ZIP format.
What's Next
These Earth observation layers are just the beginning. Combined with the IoT sensor data from our LoRaWAN network and the 1.49 million species occurrence records from GBIF, they start to paint a rich, data-driven picture of the Isle of Man's environment.
The Biosphere Observatory now covers species, marine life, tree cover, soil conditions, land temperature, weather, and live sensor data — all in one place. Every dataset is open, every source is documented, and every analysis can be verified.
Explore the new pages: