Space4Nature: Satellites, AI… and Hand Lenses

Friday 12th April 2024

A guest blog by Surrey Wildlife Trust, originally written for Surrey Nature Magazine, Issue 188, Spring 2024.

As the pioneering Space4Nature project reaches its halfway point, we check in with some of its key contributors.

Adder © Danny Green 2020VISIONFunded by a £1.2 million grant made possible by players of People’s Postcode Lottery, Space4Nature is a partnership between Surrey Wildlife Trust, the University of Surrey, Buglife and the Painshill Park Trust, which aims to map Surrey’s habitats more accurately than ever before. Using the latest satellite earth observation imagery and artificial intelligence alongside citizen scientists on the ground, it will enable the Trust to work with landowners across the county to create, restore and connect fragmented habitat and support healthier biodiversity.

Andrew Jamieson – SWT S4N Project Manager

Andrew Jamieson SWT

This is a multi-faceted project with many moving parts and I’m extremely pleased with progress to date. The University of Surrey is focusing on mapping, observation data, and machine learning. Buglife’s B-Lines has already reached its overall project target for habitat restoration. And here at SWT our citizen science programme has been fantastic. We’ve trained dozens of volunteers on survey methodology and our S4N app, and we’ve completed many surveys.

We’ve been adapting as we go. Our original bid presented an eye-catching idea, but we had many details to iron out. Some of my major responsibilities have been to agree the overall project plan and define the right outcomes and KPIs (key performance indicators that measures progress over time for a specific objective).

Taking a partnership approach naturally means we have all learned to accommodate others’ ways of working, from academic to entrepreneurial. Different organisations have different styles, timelines, and stakeholders.

I’m also struck by the level of external interest, including from other partnerships and public bodies. We talk regularly to other Wildlife Trusts, NGOs, National Parks and so on. The word is spreading.

The project’s wider effects are also potentially significant. It has boosted our own research and monitoring team and promoted the use

Andonis Blue will benefit from greater connectivity © Tom Hibbert
Andonis Blue will benefit from greater connectivity © Tom Hibbert

of technology. However, while many people are using satellite technology in conservation, for me citizen science is the key difference that can deliver accurate mapping and aid landscape recovery and connectivity.

Meanwhile, we are also working with The Land App, which provides easy-to-use technology for landowners and managers to map their land. It’s a gateway to a huge constituency of users, from farmers to water companies, who already use it to plan interventions and access funding.

Dan Banks – SWT Citizen Science Officer

Dan Banks SWTSometimes I meet people who are deterred from volunteering with SWT because they think it may be too physically challenging. This project is a great example of how we welcome everyone, whatever their age, physical strength, or mobility – we currently have more than 100 taking part – and every single hour helps.

Training is free and after three hours they’re citizen scientists, ready to go out and survey some of the most beautiful places in Surrey, such as Chobham Common, Quarry Hangers and Puttenham Common. This year we aim to expand into other areas, including urban environments.

Each volunteer is assigned a designated area. Within a one-metre quadrat they use a simple smartphone app to identify plants and record the amount of bare ground, sward height, other interesting species, signs of grazing and so on. Cheat sheets and QR codes make the whole process really easy. For accuracy we need the plants to be in flower, so most of the work is done in spring and summer. Some people do it in pairs or groups, while others prefer to go solo. We have teenagers, pensioners and everyone in between.

Our volunteers are so positive and enthusiastic, it’s wonderful. Forty-two of them came to our open forum event last summer to hear the S4N partners explain their roles. The University of Surrey presentation on satellites and data feeds was particularly popular.

It’s also fantastic to see them developing their skills and becoming recorders of more detailed wildlife information. We’re delighted to get closer to our target of one in four people taking action for nature.

Ana Andries – Lecturer and Research Fellow at the Centre for Environment and Sustainability, University of Surrey

Dr Andries assessing satellite imagery
Dr Andries assessing satellite imagery

My role is primarily technical: to combine the power of artificial intelligence and machine learning with satellite imagery, so we can rapidly identify and match habitats on a large scale. However, to make this work, the ‘ground truth’ that comes from citizen science is priceless.

Essentially, the AI learns to recognise different habitats by understanding patterns of spectral reflectance in the satellite imagery, which we match to observations on the ground. Together this data helps us efficiently create accurate habitat maps. To date we’ve made promising progress on chalk grassland and heathland distribution – and we’re continually refining the models to make them more accurate.

So far we have been making predictions about habitat quantity; next we will look at its quality, which requires a new methodology. We need to collect data from more sites to better understand the variables. Combined

with quantitative data, this will feed into our recognition and capture of biodiversity net gain (BNG), which property developers are now obliged to deliver as a condition of planning permission.

Already we know this is the most accurate mapping method we’ve ever had – and one that DEFRA and ecologists can rely on.

About the satellites

Space4Nature uses images from the PlanetScope mission launched by Planet Labs, part of the European Space Agency. The mission comprises more than 430 different satellites, which circle the Earth every 90 minutes at relatively low orbit (c.500km). The satellites each weigh less than 6kg and measure 10x10x30cm. Together they can image almost all the land on Earth every day. They use spectral reflectance to remotely identify and measure different types of vegetation. Planet Labs provides this data free of charge to S4N project, as it is for academic and research purposes.

For more information visit 

Mike Waite – SWT’s Director of Research and Monitoring

Mike Waite SWT

Having been involved in selecting our test sites, I’m glad that we haven’t had many unexpected results so far – although the Earth Search bioblitz at Painshill Park last summer was even better than we hoped, with some 300 species identified. In general, we chose heathland and chalk grassland sites that proved to be representative of their respective habitats.

The ultimate prize is an accurate understanding of habitat quantity and distribution right across Surrey. Our current picture is flawed and S4N offers a new way to remotely interpret data and provide information that is essential for connectivity modelling.

At this early stage we’re still seeing some anomalies, when the technology misrecognises a particular type of grassland, but this only helps us refine it. There are many factors to consider, such as weather conditions and time of year. In some cases, we are trying to find the sweet spot – the narrow window when a specific habitat is most easily identified.

Land managed by SWT, National Trust and other conservation organisations is already well documented, which is why we use it as test sites. The critical shift is to extrapolate onto private estates, where there are huge opportunities to improve things for wildlife.

Tawny Owl and Dormouse © Dale Sutton/ 2020VISION

These landowners tend to be receptive to the idea of acting for nature, as long as you can show them the data. With environmental policy and subsidies still in flux after Brexit, this is the ideal time to advocate with them.

As for improving the prospects of particular species, I could name three flagships that will certainly benefit from our work: Dormouse for woodland; Adonis Blue Butterfly for chalk grassland; and Adder for heathland and acid grassland. All three tend to be on fragmented sites, with a limited dispersal range, so better connectivity will be vital for their prospects.


Louis Harrington-Edmans – Space4Nature Conservation Officer, Buglife

Buglife’s practical conservation work in the UK includes B-Lines, a beautiful solution to the problem of the loss of flowers and pollinators through habitat fragmentation. It involves restoring and creating 3km-wide lines of wildflower-rich habitat through countryside and towns to provide connectivity for invertebrates and other wildlife.

Being part of S4N has enabled us to deliver this for the first time in Surrey – and the results are spectacular. Halfway through the project, we have already beaten our target of 30ha of new, restored, or enhanced B-Line habitat. From now on, we aim to keep up the momentum, reach out to more landowners, and refine our site selection. The machine-learning feedback loop inherent in S4N helps us monitor the impact of our efforts.

As the project continues, we’re learning how to address a range of different sites, from nature reserves to community green spaces, and to work with private landowners, such as Painshill Park Trust, conservation NGOs, and local authorities. For example, we’ve recently overseen reseeding at Tyting Farm near Guildford and scrub clearance at Quarry Hangers.

Above all, working with our S4N partners is a wonderful opportunity to learn. We don’t always agree on everything, but we have the same overarching goal: to improve Surrey’s habitats for nature.

Bridging the divide

A research project is uniting the academic and the practical.

Ben Siggery SWT

Ben Siggery is GIS, Research and Monitoring Manager at SWT. He is also studying for a Practitioner Doctorate in Sustainability in the Centre for Environment and Sustainability at the University of Surrey, funded by the Space4Nature (S4N) project. Here he explains how the academic and professional aspects of his work fits together.

I’m now one year into my PhD, which aims to support S4N by providing a historic baseline against which we can assess the current state of the environment. In other words, I’m trying to determine how rich life in Surrey should be if it weren’t for the damage inflicted by human activity. This lies within a field of science known as palaeoecology, or the study of past ecosystems and environments.

As a Doctoral Practitioner, I want my research to benefit practising conservationists. This is because, while paleoecology has an impressive history, it can feel inaccessible. In my PhD I am exploring the barriers to integration in conservation, experimenting with ways to overcome them and providing guidelines for academics, so they can make their work more easily translatable into effective action on the ground.

Opening the door

The first thing I learnt was that a PhD requires a vast amount of reading! Before I could start my own practical research, I needed to know what had been done already, where the gaps are, and what’s achievable. I concluded that paleoecology has huge potential to support conservation but there’s a clear disconnect between academics and professionals, and valuable information is often hidden in specialist journals. What’s more, research projects are regularly designed without the input of practical conservationists.

Compare this with medicine, which has well established and funded procedures for researchers to circulate information to practitioners. Through S4N we are attempting to emulate this culture of knowledge exchange. I’m submitting my research outputs to both academic and practitioner publications and looking at how we can reframe information that has already been published, but not widely read.

Primary Research

In the second six months of my PhD, I published a study based on the responses of 153 conservation practitioners, and their familiarity with and attitudes to paleoecology. It revealed problems such as lack of money and resources to carry out research, and poor communication with academia. However, I also identified potential solutions, including creating a centralised online database, more access to applied case studies, and better links to legislation and policy decisions.

The other recommendations to come out of the work are that academics should aim to use examples that are relevant to practitioners, co-author with experts from NGOs, provide open access to published findings, and write summaries in simple English.

Palaeoecology in action

After a lot of time in the library and at the computer, my first outing in the field was to Chobham Common. Researchers from University College London and I collected two sediment cores from ponds that have existed on the Common for hundreds of years. We pushed a long tube as deep as possible into the sediment at the bottom of the pond and extracted a sample. We then cut and bagged 1cm slices for laboratory analysis of four indicators of environmental change:

1 – Diatoms, a type of microscopic algae that only survive in very narrow ecological niches. Through them we can examine a range of variables, such as pH, water temperature, and salinity.

2 – Spheroidal carbonaceous particles (SCPs) are only produced by burning fossil fuels. Their first (oldest) occurrence in a sediment core will coincide with the industrial revolution. Increase over time can also indicate the impact of periods such as the 1950s, where there was a dramatic increase in the use of fossil fuels.

3 – Macrofossils, which are usually the remains of plants and invertebrates, can tell us about the local ecology and species assemblages of the area over time.

4 – Macro-charcoal remains tell us about the frequency and intensity of wildfires over the history of the site.

Work to do

This summer the preliminary test results should begin to reveal some of the history of Chobham Common. In due course we will see what impact our findings may have on assumptions such as SSSI designations, and whether they inform our future heathland management strategies.

Meanwhile, I will be focussing on developing another case study with our Natural Flood Management Project at Wishmoor Stream. I’ll also be experimenting with ways to improve the accessibility of paleoecology for conservation practitioners, through approaches like “recycling” existing data and looking into whether cost-effective, lower resource studies can still provide the necessary level of information for conservation work.

It looks like another busy year!