Data Challenges & Emerging Solutions

Data Challenges and Emerging Solutions

The Nature of the Biodiversity Data Gap

Biodiversity data suffers from three interrelated problems that compound its difficulty for corporate users. The first is spatial bias: the vast majority of species occurrence records in global databases (such as GBIF, the Global Biodiversity Information Facility) come from Europe and North America, leaving tropical regions, which contain the majority of global species richness, poorly sampled. A company assessing its biodiversity exposure in Southeast Asia or sub-Saharan Africa will find far less baseline data than a company operating in Western Europe.

The second problem is temporal resolution: most biodiversity survey data is collected periodically rather than continuously, making it difficult to detect rapid changes or establish robust trend lines. The third is taxonomic bias: vertebrates, particularly birds and mammals, are vastly better documented than invertebrates, fungi, or soil organisms, even though the latter groups are often more functionally important for ecosystem services such as decomposition, nutrient cycling, and pollination.

Environmental DNA (eDNA) Monitoring

Environmental DNA represents one of the most promising technological advances in biodiversity monitoring. The technique involves collecting environmental samples (water, soil, or air) and sequencing the genetic material shed by organisms - through skin cells, feces, mucus, or spores - that are present in that environment. This genetic "fingerprint" can be matched against reference databases to identify which species were present without requiring visual observation or physical capture.

eDNA monitoring can survey aquatic biodiversity in a water sample collected in minutes, whereas traditional methods might require days of electrofishing or netting surveys. Terrestrial eDNA from soil samples can reveal fungal, microbial, and invertebrate communities that are almost impossible to survey by conventional methods. For companies needing site-level baseline data in remote or difficult-to-access sourcing regions, eDNA offers a step-change in what is practically achievable. Costs have fallen dramatically as DNA sequencing technology has matured, with comprehensive eDNA surveys now feasible for a few hundred dollars per site.

Remote Sensing and Satellite Technology

Satellite-based remote sensing has transformed the ability to monitor land cover change, vegetation condition, and ecosystem extent at global scale. Several capabilities are particularly relevant to corporate biodiversity assessment:

• Deforestation detection: Platforms such as Global Forest Watch use Landsat and Sentinel imagery to detect and alert on forest cover loss with near-real-time monitoring capability. This enables both companies and investors to monitor supplier landscapes continuously rather than relying on annual snapshots.
• Habitat condition assessment: Indices derived from satellite imagery, such as Normalised Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), proxy vegetation health and productivity. More advanced analysis using hyperspectral imagery can estimate species diversity and detect invasive species.
• Mangrove and wetland mapping: Radar-based satellites can penetrate vegetation canopy and map mangrove extent and density with high accuracy, enabling continuous monitoring of these critical coastal ecosystems.
• Coral reef monitoring: High-resolution satellite imagery and drone surveys are enabling global coral reef extent and condition mapping at scales previously impossible.

Artificial Intelligence for Species Identification

Machine learning models trained on large image databases are now capable of identifying plant and animal species from photographs with accuracy rivalling expert taxonomists in many groups. Apps such as iNaturalist, Pl@ntNet, and Merlin Bird ID have democratised species identification, enabling non-expert observers to generate and submit accurate biodiversity occurrence records at scale.

For companies, AI-assisted species identification has several practical applications. Automated acoustic monitoring devices can record and identify bird calls, bat echolocation, and insect sounds continuously in the field, building up longitudinal species abundance datasets with minimal labour. Camera trap networks with AI-assisted image processing can monitor mammal populations across vast landscapes with a fraction of the traditional survey cost. In the marine environment, AI analysis of underwater audio recordings can identify cetacean, fish, and invertebrate species from acoustic signatures.

Citizen Science and Crowd-Sourced Data

Citizen science platforms have dramatically expanded the geographic coverage of biodiversity occurrence data. GBIF aggregates data from hundreds of partner institutions and platforms, with iNaturalist alone contributing over 160 million observations by 2024. While citizen science data has quality control challenges, particularly for rare or difficult-to-identify species, it provides valuable presence/absence records across geographies that would be prohibitively expensive to survey professionally.

For companies, citizen science data can serve several functions in a LEAP assessment: providing baseline species occurrence data for screening priority locations, detecting invasive species before they become established, and engaging employees and local communities in biodiversity monitoring programmes as part of a broader nature stewardship commitment. Several companies have developed employee citizen science programmes that combine internal engagement with genuinely useful ecological data collection.

Spatial Finance: Integrating Nature Data into Financial Decision-Making

Spatial finance refers to the integration of geospatial data about physical assets and their environmental context into financial analysis and decision-making. It represents the convergence of satellite earth observation, biodiversity data science, and financial risk management into a new analytical discipline.

For financial institutions, spatial finance tools enable the physical location of financed assets (factories, farms, ports, mines) to be systematically overlaid with spatial biodiversity risk data. This allows banks and investors to identify which loans or investments carry elevated nature-related physical risk based on the ecological context of the assets they finance, not just on borrower self-disclosure. The Oxford Sustainable Finance Programme and Satellite Applications Catapult have been instrumental in developing and promoting spatial finance approaches, and the TNFD explicitly references spatial finance as a tool for financial institutions implementing LEAP.

The Emerging Global Biodiversity Monitoring Framework

The Kunming-Montreal Global Biodiversity Framework includes a monitoring framework that specifies headline, binary, and component indicators for tracking progress against each of the 23 targets. This policy demand is accelerating investment in global biodiversity monitoring infrastructure, including expanded satellite capabilities, standardised field monitoring protocols, and digital data-sharing platforms.

GBIF, the primary global repository for biodiversity occurrence data, has committed to expanding its data partnership network and improving data quality verification in response to KM-GBF monitoring needs. The IUCN Red List assessment programme is accelerating species assessment coverage to provide more current threat status data. These improvements at the public data infrastructure level will progressively address the data gaps that currently constrain corporate biodiversity assessment, making TNFD-aligned reporting more achievable and credible over time.