Conserving biodiversity

Mapping the urban tree canopy using artificial intelligence.

The accelerated rate of biodiversity decline due to climate change is alarming.  We are all part of nature and need to work with it and not against it. Declining habits is a key issue which can be addressed in part by increasing trees and woodlands, particularly in urban spaces.

Forth-ERA objectives

  • Map individual urban trees and woodlands
  • Determine how tree cover changes locally (for example with SMID)

The University of Stirling are harnessing artificial intelligence (convolutional neural network) to detect trees in aerial imagery. A web-based application assists with data visualisation with an accuracy of around 92%.

We are using high performance computing to process data and perform complex calculations at high speeds (at quadrillions of calculations per second).  Data is bought to life through data visualisation in a secure, scalable cloud-based data centre. Customisable dashboards drill into pertinent data for the end-user.

Our work in the area compliments perfectly our aims with the Forth Climate Forest.

A living digital laboratory. Dr Adam Varley, Forth-ERA Research Fellow in Data Analytics on how Forth-ERA is harnessing the digital revolution to understand environmental processes.

Application of data measurement and analysis to support biodiversity and natural capital

Working with NatureScot, we have identified a number of areas that Forth ERA could support Scotland’s nature-based priorities. For example, these include effective monitoring of peatland restoration, coastal stability and salt marshes and remote monitoring of water bodies. These programmes would support a wide range of regional climate priorities, and critically provides data to value natural assets and observe a return on investment on restoration projects.

Contact us

Get in touch with us at forth-era@stir.ac.uk to discuss how our consultancy expertise, teamed with technology, artificial intelligence and data analysis can help your organisation.