Space4Nature: Empowering Nature Recovery with Citizen Science and Earth Observation

Space4Nature was established in 2022 to address a biodiversity challenge faced by conservation organisations in the UK county of Surrey (South-East England): they need up-to-date fine-scale habitat mapping to guide restoration, monitoring and nature-recovery planning. However, traditional field surveys alone are expensive, time-consuming, difficult to repeat regularly, and often constrained by access to privately owned land. To address this, the University of Surrey (UoS), Surrey Wildlife Trust (SWT) and Buglife developed a collaborative approach that combined Earth Observation (EO) satellite data, ecological expertise, structured citizen science and machine learning (ML).

Using standardised surveys designed around UK Habitat Classification criteria (Level 4 and 5), over 350 trained volunteers (citizen scientists) collected data from over 1000 quadrats, capturing habitat data for chalk, acid and neutral grasslands, as well as heathland, across 11 test sites, between 2022 and 2025. These field observations were combined with PlanetScope satellite imagery, a range of relevant vegetation indices, as well as topographic parameters and soil type data to produce 3 m resolution habitat maps with high internal classification performance, alongside independent validation across many long-term monitored grassland sites.

The resulting maps reveal the fragmented distribution of the chosen habitats, providing an evidence base for restoration targeting habitat inventory and Local Nature Recovery Strategy-related planning.

The project’s main current user bodies are SWT and Buglife, but the methodology is intended to be reproducible for other Wildlife Trusts, environmental NGOs, local authorities and wider actors involved in local nature-recovery planning.

Conservation organisations in Surrey need more than general habitat statistics; they need spatially detailed, practical evidence showing where priority habitats remain, how fragmented they are, and where restoration or further survey effort would have the greatest value. This need is especially important for habitats such as chalk and acid grassland and lowland heathland, which persist as small, scattered and often isolated patches across the county and are thus vulnerable to degradation. For organisations working on habitat recovery, connectivity, and long-term stewardship, broad national habitat products are often too coarse, too generalised, or too quickly outdated to guide action at site and landscape scales. The project, therefore, addressed a real information gap between the level of detail required by local conservation practice and what conventional mapping approaches could realistically provide.

A second challenge is that expert-led surveys, while essential, are resource intensive and difficult to scale, depending on staff time, repeat visits, funding, and access to land. Space4Nature responded to these constraints by combining ecological expertise with structured citizen science and Earth observation, so that evidence collection could be expanded without losing ecological rigour.

The main current users are Surrey Wildlife Trust (SWT) and Buglife. SWT needed better habitat evidence to support conservation planning, habitat inventory, connectivity assessment and engagement with land managers, while Buglife needed outputs that could help identify opportunities for habitat restoration and creation. More broadly, the approach will be provided to other Wildlife Trusts across the UK, environmental NGOs, local authorities and statutory actors involved in local nature recovery planning, including the evidence needs associated with Local Nature Recovery Strategies.

A key product from the work is the app-supported field evidence generated through the Survey123-based citizen science workflow. The app was designed to guide volunteers through standardised habitat surveys using UKHab-informed questions on indicator species, vegetation structure, bare ground, scrub, management, and site condition. This meant that the field records were not just lists of species observations in isolation, but structured habitat information that could be linked directly to mapped outputs. For users, this created a more transparent evidence chain between field observations, ecological interpretation and final habitat products.

Rather than asking users to interpret complex EO datasets directly, the Space4Nature methodology translates them into practical habitat evidence that can be used in day-to-day conservation planning and prioritisation. The main products are fine-scale habitat maps at 3 m resolution, produced for priority habitats such as chalk and acid grassland and lowland heathland across Surrey.

The project also generated decision-support layers and validated outputs that are easier for users to interpret. These include mapped patterns of habitat extent, fragmentation and potential restoration opportunity, helping SWT and Buglife identify habitat stepping stones, target restoration and habitat creation, and support wider planning discussions.

Space4Nature is built on a combination of EO derived data, environmental spatial data, structured citizen-science field surveys, and expert ecological validation. Across the three habitat types, the core EO input was PlanetScope imagery in 8 spectral bands useful to derive critical vegetation indices and 3 m spatial resolution, which is fine enough to detect small, fragmented habitat patches that are often missed or blurred in coarser national products. Multi-temporal imagery (May-September) was used so that habitat mapping could reflect seasonal vegetation characteristics rather than relying on a single snapshot in time. This was particularly important for distinguishing habitats that may look similar on one date but differ in structure, composition or seasonal behaviour. For example, the spectral signature of acid grassland is often confused with heathland vegetation.

These satellite data were combined with a set of environmental variables that help explain where habitats are likely to occur. These included topographic variables derived from LiDAR-based terrain models, such as slope, aspect and plan and profile curvature, topographic wetness index, etc, together with soil information from national soil datasets. These layers were important because fragmented habitats such as chalk and acid grassland are not determined by vegetation signal alone: they are also shaped by underlying ecological conditions such as soil type, drainage, exposure and landform. Therefore, when combining EO with these environmental covariates, the project generated habitat evidence that was ecologically more realistic and easier to interpret in management terms.

A defining feature of Space4Nature was its use of structured citizen-science observations collected through 1 m² quadrat surveys using the ArcGIS Survey123 app. The app was designed around UK Habitat Classification (UKHab) criteria and guided volunteers through a standardised set of habitat questions, including indicator species, vegetation structure, bare ground, scrub cover, signs of management and condition. This meant that the field data were collected in a consistent format that could be linked directly to habitat modelling. Citizen scientists were recruited and trained by SWT, and the survey design was carefully tested to make it accessible while retaining ecological value. The process also provided participants with learning and experience.

Ecological expertise from SWT was central throughout. SWT helped select sites, co-designed the survey protocols, trained volunteers, reviewed submitted records, and validated whether quadrats met the relevant UKHab habitat criteria before data were used in the modelling. In this way, citizen science did not replace expert ecology but worked within a quality-controlled framework supported by digital tools, field methods and expert judgement. The result was a robust evidence chain linking satellite observations, field data and ecological validation into habitat products that conservation users could trust.

Space4Nature has delivered impact in three connected ways by improving the evidence available for conservation action, by widening public participation in nature recovery and by demonstrating a model that can be transferred and built on beyond the original project. Together, these benefits matter because they help elevate habitat mapping from a one-off technical exercise towards a more practical, inclusive and repeatable form of environmental decision support.

First, the project has improved the evidence base for conservation action in Surrey. The fine-scale habitat maps produced through Space4Nature have filled important gaps left by broader national datasets and by field surveys that are difficult to repeat at the county scale. For users such as SWT and Buglife, this has meant better spatial evidence on habitat extent, fragmentation and likely restoration opportunities. The outputs have supported more rapid grassland habitat inventory, better targeting of restoration and habitat creation, and a clearer understanding of where ecological connectivity could be strengthened. For Buglife, the Space4Nature outputs are particularly relevant to the Surrey Chalk Lines project, which focuses on calcareous grassland restoration through advice to landowners and managers, community engagement, botanical and entomological surveys, and practical habitat management work. The mapped evidence will help Buglife identify where chalk grassland fragments remain, where habitat creation could strengthen ecological links, and where restoration activity could contribute most effectively to pollinator recovery across the Surrey Hills landscape.

Second, Space4Nature has shown that citizen science can generate value well beyond data collection alone. The project created opportunities for volunteers to learn plant identification, habitat survey methods and the wider purpose of conservation mapping, helping more people engage directly with nature recovery in Surrey. SWT built a local citizen-science network around this work and created a feedback process so that volunteers could see how their surveys contributed to real outputs and decisions. This strengthened community stewardship and made the project more socially grounded, while also broadening the evidence base available to conservation organisations. It also provided a strong foundation from which SWT could develop other citizen science projects and volunteer engagement activities.

Third, the project provides a scalable and reproducible approach that is relevant beyond a single habitat or organisation. The project provides a methodology as a transparent and accessible framework for strategic conservation planning and is intended to be relevant for other Wildlife Trusts, environmental NGOs, local authorities and local nature recovery actors. One important legacy strand is now extending the approach to small ponds and wetlands through PhD research that uses freely available EO and machine learning, with future plans to incorporate citizen science for finer UKHab wetland classification.