Natural England removes clouds using machine learning

The project, run by Natural England, uses satellite imagery, geospatial information, and field data records to create a national map of England’s habitat. Living England uses AI instead of manual surveys to track habitat changes more efficiently. The government says this will speed up planning and land-use decisions while better protecting nature.

Living England shows how combining Machine learning is improved by using multiple datasets and human observations. These datasets include satellite data collected by Copernicus, a component of the European Union space programme that focuses on Earth observation. The Natural Capital and Ecosystem Assessment programme (NCEA), which was launched in 2016, is funded through the Department for the Environment, Farming and Rural Affairs.

Copernicus Sentinel-2, a joint programme of the European Space Agency (ESA), and European Union (EU), provides imagery of Earth’s surface with a spatial-resolution ranging from 10m to a 60m. As an optical satellite the land surface can be obscured by clouds and its appearance altered by cloud shadow. Both of these factors reduce the quality of imagery used for habitat modelling.

Clouds and cloud shadows need to be removed before modelling to produce cloud free mosaic images of the surface to use in modelling. To accomplish this, data collected on the ground by surveyors is correlated with satellite images from the Sentinel-2 program using machine learning techniques in order to identify clouds. Living England also uses the Environment Agency’s National LIDAR Program, which provides topographic data with a high spatial resolution (1m). Other datasets include climatic and soils data from the British Geographical Survey’s HadUK gridded product, geology, and soils data from Cranfield NATMAP soilscapes and Ordnance Survey. Our Living England project harnesses the power of AI in order to inform and support decisions much more efficiently.

Sallie Baily, Natural England.

Natural England’s chief scientists Sallie Bailey commented on the project. “Nature Restoration, Development and Economic Growth are not opposing Forces – they can work together to create sustainable futures for people and wildlife.”

Our Living England project is harnessing AI to inform and help support planning decisions more efficiently. This will allow us to make the greatest impact on nature recovery and deliver the new homes, infrastructure and services the country needs.

In order to support its ambition to transform the public services using AI, the government has published a AI Playbook.

Peter Kyle, Technology Secretary, said: “The publication today of our AI Playbookcomes with a call-to-arms for tech specialists in the public sector. Use the guidance we’re sharing to put AI to use in your organisations whiplash-fast, so we can fix our broken public services.” AI is used to improve weather forecasts and maintain high standards at MOT test centres.