Ground Truth:
Why Satellites Aren’t Enough
Behind The Tech: The Synergy between NatureOS & Satellite Data
But… is this enough?
As with all technology, there are limitations that have yet to be overcome. Carbon stocks often get inaccurately represented by satellite data due to tree canopy cover. As leaves and branches form dense canopy cover more than a hundred metres off the forest floor, lasers used in LiDAR capturing are unable to uncover vegetation and terrain hidden underneath.
Satellite data alone also cannot account for local variations in factors like tree species composition, wood density, and site conditions, which can significantly influence carbon stocks.
With such huge gaps in knowledge from relying on remote sensing technology alone, the need for a more reliable method for measurement is evident.
Ground Truth: The Critical Piece
Ground truth serves as the anchor that grounds remote sensing data in reality. By directly collecting measurements and observations on-site, scientists can validate and cross-reference new data with satellite and LiDAR-derived data, ensuring accuracy and reliability. This includes collecting field measurements of tree diameter, biomass, soil carbon content, plant species, water chemistry and more. With boots on the ground, we use ground truth to verify and eliminate our biases.
While it is not convenient at all to measure ground truth, the importance of acquiring accurate carbon estimations for land projects and unlocking nature investments cannot be understated. The critical problem here is that ground truth collection is manual: a laboursome task that involves many days of field measurements, sampling and plot-based monitoring.
An Uphill Task
Ground truth collection for forest carbon biomass typically hinges on manual methodologies, a process steeped in traditional techniques, especially in Asia:
-
This approach entails employing tools like measuring tape and clinometers to document parameters such as height and diameter at breast height (DBH), standardised at 1.30 meters
-
These raw field measurements serve as the bedrock for subsequent calculations employing allometric models, facilitating biomass estimation
It’s no wonder, then, that ground truth collection is time consuming, labour-intensive, and unfortunately prone to human error. Biomass estimates may thus vary wildly.
The Arkadiah Difference
This is where Arkadiah’s expertise with powerful imaging technologies makes a difference. The Terrestrial LiDAR System (TLS), for example, captures high definition digital twin of trees and forests, gathering critical data including diameter, height, crown radius and geotagged position of trees. The carbon estimation, based on tree volume, is then calculated from 3D point clouds.
Other methods are available to capture data for huge areas of land — through drones and unmanned airborne vehicles (UAVs), or airborne LiDAR at a national scale.
Collecting Data 10X Faster, with LiDAR
LiDAR’s sheer power at imaging larger swathes of forests at one go means that what used to be done manually over 6 days can now be done in just half a day. Digitised data collection has never been more optimised — a tenfold productivity in biomass estimation is testament to that.
Carving a Path Led by Experts
An ambitious venture like Arkadiah’s is impossible without the passion and deep domain knowledge of its leading experts. Gerry Ong and Dr Deepthi Chimalakonda are two incredibly valued members of our team who are helping to make efficient and traceable mass biomass estimation a reality.
Gerry Ong is Arkadiah’s Head of Geospatial and also one of its co-founders, bringing more than 30 years of experience to the table in geospatial technology. Having worked on a large national plan to map large national plan to map tropical forest inventory, Gerry leads the data collection necessary to reconstruct digital twins of trees and forests.
Dr Deepthi is Arkadiah’s Head of Carbon & Biodiversity, and is our resident ecologist who brings 15 years of experience in biodiversity assessments and ecological modelling.
Working with Local Communities
We work closely with local partners in target regions to gather data, instead of sending large teams to conduct surveys. Our local partners have in-depth knowledge of their local ecosystem and terrain, opening a wealth of information about the location of study.
In Brunei, our collaboration with a local forestry expert and his team offers invaluable insights into the ecology of peat swamp forests, including the local ecology, hydrology and ecosystem dynamics, and the challenges they face in the field. We were also knee-deep in the cultural context of the area and its effect on project activities.
In Thailand, we conduct ground truth activities in partnership with organisations that prioritise investing in and engaging local communities in environmental protection efforts. While our scientific methodologies pertaining to forest carbon stock estimation may differ, we prioritise understanding both the distinctions and commonalities in the metrics achieved through various approaches resulting in knowledge exchange and capacity building.
Our local partners play key roles in facilitating communication in local languages and navigating cultural nuances, ensuring clear engagement with community members. They also assist in navigating complex regulatory frameworks and compliance requirements specific to the project location, reducing risks and ensuring legal adherence.
Arkadiah emphasises that working closely with local partners builds community trust, encourages participation, and fosters a sense of ownership crucial for the project’s long-term sustainability and success.
The Magic of Synergy: NatureOS & Satellite Data
Our work, incredibly, does not stop there. We integrate ground truth measurements from sample plots with our satellite data in NatureOS to enhance the accuracy of biomass data.
Leveraging sophisticated algorithms and honest, boots-on-the-ground work using ArkVantage, we compared the Aboveground Biomass (AGB) of 2 different forested sites in Southeast Asia. Consistently, our ground truth-adjusted model prevails in terms of capturing biomass that would not be detected by satellites alone.
As a repeatable observation, our novel approach effectively corrects AGB underestimations inherent in solely satellite-derived data with 10%-80% improvements based on ecosystem, GEDI footprint etc. The numbers speak for themselves.
NatureOS: Integrating Ground Truth & Remote Sensing
Our proprietary model forms a critical backbone of NatureOS, Arkadiah’s natural accounting platform to support businesses in the development and management of their nature portfolios and environmental impact.
With a comprehensive suite of tools, NatureOS captures data from ESA, GEDI and Landsat across key quality indicators and ground truth data to support project developers, land owners, investors, carbon buyers and insetters on their journey to become nature positive.
As an industry first, NatureOS also allows users to upload their ground truth data into the platform, and Arkadiah’s AI-driven, self-learning system calibrates that data to improve accuracy for biomass inventory estimation.
