Identifying climate change with Satellite Earth Observation data and artificial intelligence.
Dubai is a rapidly developing city. This growth has had a big impact on the natural environment, natural resources and native habitats.
Satellite Earth Observation (EO) data has long been an important source of information for measuring and monitoring how the environment is changing. However, producing geospatial information for large areas from EO data is time consuming and costly.
Ordnance Survey (OS) and Deimos Space UK worked with the Mohammed Bin Rashid Space Centre (MBRSC) in Dubai to automate the production of geospatial information from satellite EO data using machine learning algorithms.
The collaboration had to work to make the EO data usable to meet the needs of a range of stakeholders, including those dependent on the natural environment.
This project focused on a number of key challenges:
- Tracking the growth and health of important vegetation.
- Providing reliable data to ensure correct subsidy payments to farmers.
- Maximising data production efficiency and value from satellite EO data.
- Creating an interoperable machine-to-machine data storage capability and strategy aligned with other Dubai data initiatives.
"Our collaboration with OS was very valuable both in terms of developing technical capability and exploring new business opportunities."Michael Lawrence, Business Development Director, Deimos Space UK LTD.
As palm trees and mangroves have cultural and economic value in Dubai, Ordnance Survey and Deimos Space UK developed a prototype palm tree and mangrove feature for EO data using state-of-the-art deep learning techniques.
This supported the development of an interoperable data model to easily share data with other government departments and inform decisions.
A Spatial Data Infrastructure Strategy for MBRSC was also created to ensure data aligns with the latest developments in the Dubai Spatial Data Infrastructure and the wider Dubai geospatial sector.
The project was supported by the UK Space Agency and UK Department for Business, Energy and Industrial Strategy (BEIS) as part of the Gulf Science, Innovation and Knowledge Economy Programme.
- The automatic production of geospatial information was made possible with equivalent or greater accuracy than manual processes.
- Efficiency savings in data production (time and cost) were achieved.
- An independent, interoperable vegetation index data model allowed for more frequent data updates for better measurement and monitoring.
- Greater understanding of stakeholder requirements of MBRSC data allowed for alignment to wider Dubai initiatives such as Smart Dubai.