Space Intelligence

Those who know me well know I'm constantly banging on about forest degradation. You know, the second D in REDD - so often it feels like it's just 'Reducing Emissions from Deforestation', with Degradation just ignored. Why? I think largely it's because it's difficult to map. Deforestation is binary - the change from forest to non-forest. Easy(ish) to spot, easy to find maps and statistics, often simple to work out the drivers - even if not always easy to actually stop. But forest degradation is a 'shade-of-grey' - the removal of some trees from a forest, which then remains forest. Estimates vary as to how much land or carbon is impacted by each annually, and estimates are complicated by the fact that degradation often ultimately leads to deforestation at a later date (so both can occur to the same area at different times), but also that degraded forest often recovers. However, a scientific paper I was involved with (see picture) showed that in the woodlands of southern Africa, 5x more area was degraded than deforested in the 2010s, and about 55% of the carbon loss. And some papers have claimed that degradation might represent as much as 70% of all forest carbon losses. Even if that is an overestimate, it's clearly important enough to monitor well and try to step and reverse. The challenge is that most monitoring tools weren't built to catch it. Global alert products can be useful for large-scale deforestation — but detection thresholds designed to work everywhere miss the smaller, subtler events. We've written about how our Space Intelligence product DisturbanceTracker - which uses 10m resolution optical and radar data to perform monitoring tuned to a specific site - performs differently, detecting both deforestation and degradation. We show some comparison data from Kalimantan and Gabon shows, and why the timing of detection matters almost as much as the detection itself. If you're working on REDD+ projects or forest carbon portfolios, I think this is worth reading - see comments for the link.

Most deforestation alerts are designed to catch the obvious #deforestation: large clearings and forest that's gone overnight. Selective logging doesn't look like that. Trees are removed, canopy closes over, and the forest appears intact — even as carbon and biodiversity erode underneath. According to recent research, 69% of net tropical forest carbon loss comes from small-scale degradation, not outright clearance. Standard alert products, built to work globally, often miss it. In our latest piece, we look at why global alerts struggle with this type of disturbance, what SAR-based near real-time monitoring detects instead, and what that difference means for #REDD projects and carbon portfolios. Link in comments. #forestmonitoring #DisturbanceTracker

Reminder that we're going live next Thursday, May 14 at 2pm BST to show you how DisturbanceTracker monitors small-scale degradation in REDD projects.   If you're relying on open-source alerts or annual reports for your REDD+ projects, you're likely missing small-scale selective logging until it's too late.   During the 30-minute demo, Dr. Alexis Moyer and prof Ed Mitchard will compare DisturbanceTracker side-by-side with open-source tools across sites in Brazil and Indonesia. You'll see exactly how early alerts can help you act and prevent further degradation quicker.   Save your spot here: https://lnkd.in/e-jJKx9V   (As always, if the timing doesn't work, register anyway and we'll send a recording)  #redd #disturbancetracker #deforestation

Verra's REDD+ project baselines have been a big part of my life the past six years of so. I think people often think that Verra's bringing of baselines in house under what was first called the "Consolidated REDD+ Methodology", and is now the catchy "VM0048", was a response to articles critical of REDD+ baselines published in 2023. But I first became aware of (and started inputting on) the methodology back in 2020, and in 2021 we at Space Intelligence first tested the creation of jurisdiction-wide Forest Cover Benchmark Maps to create 'Activity Data' (well calibrated estimates of historical deforestation used to make risk maps), for Tanzania. It was a big challenge because the accuracy requirements for these country-wide maps were stringent, and dry forest areas are hard to map really well! But overall, we found the approach worked: we could make consistent jurisdiction-wide datasets, that could form the baseline for any REDD+ project in that jurisdiction. The methodology went live in 2023, and there was a massive challenge in terms of the number of benchmark maps Verra needed to create to meet project developer demand. They ended up procuring these from a large number of companies, including us, and we went on to produce forest benchmark maps for six more countries, from Argentina to Timor Leste via Ethiopia, Indonesia, Kenya and the Republic of Congo. As projects start actually generating VM0048 credits, and some of the initial risk maps reach the end of their validity period, Verra are moving to a more mature and long-term contracting phase to continue receiving these maps from providers where needed. I'm delighted that we are one of the chosen providers for this next phase, and I look forward to making many more of these maps - and them being used as baselines to enable projects to sell carbon credits and protect forests!

Verra has selected Space Intelligence as just one of 3 data service providers to deliver the allocated risk maps for VM0048 and its module VMD0055. We’ll produce jurisdictional activity data, forest cover benchmark maps and allocated deforestation risk maps over a 5-year period. The first forest-rich jurisdictions are: 🇵🇭 The Philippines - allocated deforestation risk map. 🇧🇷 Mato Grosso State, Brazil - activity data, forest cover benchmark map, and risk map. Concentrating the work with 3 proven DSPs allows Verra to increase efficiency of the data development process - giving project developers and offtakers the clarity they need to plan baselines, supply, and timing. Space Intelligence already produced 7 of the first Forest Cover Benchmark maps for Verra (Tanzania, Republic of Congo, Kenya, Ethiopia, Indonesia, Timor Leste & Argentina), so this acts as an extension of this partnership. From prof. Ed Mitchard, Chief Scientist and Co-Founder: "This new methodology achieves the vision we set out in 2020 - a modern approach building on lessons from the first generation of REDD+ methodologies, with robust and conservative baselines built by independent providers like Space Intelligence. We're excited to expand our partnership with Verra and be one of the partners in creating these over many more countries over the coming years." Read more: https://lnkd.in/e_QTvSve #REDD #VM0048 #CarbonMarkets

Reminder that we're going live next Thursday, May 14 at 2pm BST to show you how DisturbanceTracker monitors small-scale degradation in REDD projects.   If you're relying on open-source alerts or annual reports for your REDD+ projects, you're likely missing small-scale selective logging until it's too late.   During the 30-minute demo, Dr. Alexis Moyer and prof Ed Mitchard will compare DisturbanceTracker side-by-side with open-source tools across sites in Brazil and Indonesia. You'll see exactly how early alerts can help you act and prevent further degradation quicker.   Save your spot here: https://lnkd.in/e-jJKx9V   (As always, if the timing doesn't work, register anyway and we'll send a recording)  #redd #disturbancetracker #deforestation

We are headed to Singapore! We are excited to share that our team has been selected for the Blue Catalyst Challenge, taking place in Singapore from 11-22 May 2026. Organised by Hatch Blue and WWF-Singapore, with support from the Singapore Economic Development Board (EDB), we will refine our tech solution for high-integrity blue carbon alongside world-class experts from across the global carbon ecosystem. We warmly welcome our partners and peers to join us as we share our progress - RSVP for our Demo Day on 22 May: https://luma.com/4s7j3dbi #bluecatalyst #singapore #mangroverestoration