CASE STUDY

Remotely Sensing Habitat Condition for Landscape Recovery

Discover how AtkinsRéalis and AiDASH are revolutionising the way large-scale habitat assessments are conducted through remote sensing and AI. The Eelscapes Landscape Recovery (LR) scheme, hosted by WWT and funded by Defra, faces a major challenge: meeting Defra’s LR Monitoring and Evaluation requirements for baseline habitat assessments across vast project areas (up to 61,500 ha).

Traditional methods are costly and time-consuming, but our data-driven solution using the BNGAI™ and River Restoration Studio tools offers a smarter, more efficient way to classify and assess habitat condition.

Unlock the case study & learn how BNGAI™ transformed habitat assessment for Eelscapes.

Introduction

Eelscapes is one of the flagship Round 1 Landscape Recovery (LR) pilot projects funded by Natural England through a Defra Project Development Grant. The project is being led by the Wildfowl & Wetlands Trust (WWT), in collaboration with Gloucestershire Wildlife Trust and the Environment Agency, alongside a diverse network of local land managers.

Currently in its development phase, Eelscapes is laying the groundwork for long-term ecological restoration across a complex and expansive landscape. This phase is not only critical for establishing a vision and direction for the project—it also sets the data foundations for how progress will be measured, reviewed, and scaled in the years to come.

The team’s goal: to create a comprehensive, verified habitat baseline across the entire project area—delivered efficiently, designed to be repeatable, and structured to inform future decision-making and Defra reviews.

THE CHALLENGE

Baseline habitat assessment at scale

To meet Defra’s Monitoring and Evaluation requirements during the Project Development Phase, the Eelscapes team needed to assess and classify all habitats across the project area—both by type and condition. For this, the use of Statutory Biodiversity Metric tools was non-negotiable.

  • Habitat types needed to be classified using the Statutory Biodiversity Metric framework, based on the UK Habitat Classification (UKHab).

  • Habitat condition needed to be assessed either via Defra’s habitat condition sheets or, for rivers, using the Cartographer’s River Condition Assessment (RCA).

But the scale of the task was significant. With over 1,700 hectares of land spanning multiple ownerships and habitat types, conducting purely manual field surveys would have required enormous resources. It would also have been heavily constrained by access issues, seasonal limitations, and the time pressure of a 5-month reporting window.

The question was clear: How do you deliver a compliant, reliable habitat baseline across such a large and fragmented area—quickly, affordably, and with ecological integrity?

Screenshot from the BNGAI™ platform on an external project’s habitat condition assessment

Example of River Studio outputs (AtkinsRéalis)

THE SOLUTION

A collaborative, data-driven approach

To answer this challenge, the Eelscapes team partnered with environmental technology leaders AtkinsRéalis and AiDASH. Together, they implemented an innovative, data-driven approach that combined satellite intelligence with expert ecological input.

At the heart of the approach were two cutting-edge tools:

  • BNGAI™ – AiDASH’s AI-powered platform for automated habitat classification and BNG condition assessments.

  • River Restoration Studio – AtkinsRéalis’ digital tool for assessing river condition using open datasets, survey inputs, and modelled outputs.

The result was a remotely sensed, high-resolution habitat and river condition map—cross-checked and refined by ground-truth survey data—delivered across all 1,700 hectares in less than five months.

This hybrid method provided a strategic and scalable alternative to traditional survey approaches while meeting all of Defra’s baseline requirements.

1700 HA

processed from kick-off to reporting in under five months

50+

sites split into batches and prioritised for processing

26 SITES

selected for ground surveys to validate the model

<5 Mo

from project kick-off to reporting through BNGAI™

Services provided

The process in practice

The Eelscapes team followed a structured methodology to combine automation with ecological rigour:

  1. Review of existing survey data

    The team began by gathering and assessing all available ecological data and field records to establish a starting baseline and identify data gaps.

  2. Automated habitat classification

    Using satellite imagery, AiDASH’s BNGAI™ platform generated a high-resolution habitat classification across the full 1,700 ha in alignment with the Statutory BNG Metric.

  3. Initial habitat condition estimation

    Where condition data already existed, it was integrated. Where it didn’t, the BNGAI™ Survey app was used to estimate habitat condition based on AI-driven analysis and remote data inputs.

  4. River condition analysis

    AtkinsRéalis’ River Studio assessed river habitats using a mix of open-source datasets and local field data, generating a detailed picture of baseline river health across the landscape.

  5. Survey strategy development

    A targeted ground-truthing strategy was developed in collaboration with the LR project team. Field surveys were prioritised for areas of high ecological value, high model uncertainty, or missing baseline data.

  6. Model refinement

    Data from ground surveys was used to improve the accuracy of the modelled outputs, increasing confidence in the resulting baseline.

  7. Custom project platform

    AiDASH created a bespoke, interactive version of the BNGAI™ platform for the Eelscapes team—allowing stakeholders to explore the data, interrogate outputs, and use the tool to support ongoing delivery and planning.

Key benefits

This hybrid approach delivered a step-change in how baseline assessments for landscape-scale projects can be approached:

Time- and cost-efficient

The entire 1,700 ha landscape was processed from kick-off to reporting in under five months—far faster than traditional methods.

High model accuracy

Habitat classification and condition estimations were verified and refined through ground surveys and expert review.

Season-independent

Unlike field surveys, this method isn’t limited by seasonal access or weather conditions.

Strategic survey planning

The use of modelled outputs allowed the team to focus fieldwork on high-priority areas, increasing efficiency and ecological value.

Repeatable and verifiable

The methodology is designed for reuse at future monitoring checkpoints, enabling a clear before-and-after comparison over time.

Access to remote areas

Remote sensing allowed the team to monitor sites that were otherwise difficult to access or survey.

Scientifically validated

Outputs were reviewed and validated by accredited ecologists and aquatic scientists, ensuring ecological integrity.

Stakeholder collaboration

The interactive platform served as a shared workspace for cross-organisation collaboration and transparency.

Long-term delivery support

Establishes a strong baseline for tracking habitat change and informing future funding and planning decisions.

Watch the webinar

Discover how WWT, AiDASH and AtkinsRéalis assessed 1,700 ha in under five months using remote sensing and AI.

See how BNGAI™ can help your projects

Get a demo of BNGAI™ and see how remote sensing and AI can accelerate your BNG assessments.