Nov 16, 2026

Forecasting Ecological Dynamics

Earlier this month, Climate Spheres Research Director and Lead Climate Scientist, Dr. Rosalind Haskins, attended an event entitled; "Forecasting ecological dynamics in a changing world: a critical discussion" hosted by the Royal Society, which brought together ecological modellers from across academia and industry for some wonderfully frank discussions on the state of the science. This event was particularly  fascinating from the climate science and modelling perspective and this blog shares  some thoughts on the differences between the two fields.

The crisis of biodiversity loss is increasingly being recognised as having serious implications for human wellbeing. Global frameworks such as The Biodiversity Plan (including the 30 by 30 target) are being agreed in hopes of being able to ‘bend the curve’ of biodiversity loss. However, ecological modelling remains roughly 30 years behind climate modelling. Reasons for this include that ecology is less deterministic (‘more messy’) than climate physics and that where climate science could use preexisting meteorological datasets, biodiversity modelling lacks extensive historical baseline data for validation. Ecology also requires more ‘on the ground’ labour intensive studies, while climate scientists can benefit more strongly from satellite data. Transitioning the ecological sector from passive, local, monitoring to active mitigation of biodiversity loss will require a step change in predictive capability. 

Key takeaways from the event discussions include:

  • Interspecies dynamics remain un(der)represented in ecological models. The models will need to significantly increase in complexity in order to capture complex interspecies dynamics that could greatly impact ecosystem balance, resilience, and predictability.

  • Non-ecologist coders are increasingly applying AI models to open biodiversity data, without understanding the limitations of their results. Frameworks must ensure that qualified ecologists remain central to biodiversity data analysis and model interpretation to prevent real harm from being caused.

  • The relative simplicity of ecological models has been extremely useful for enabling communication with policymakers and the general public. 

  • The human impacts of biodiversity loss are not as obvious as in the case of extreme weather events, and do not yet trigger immediately obvious business losses or major insurance liabilities. As a result, the commercial investment lags despite clear long-term risks to global supply chains.

Climate science demonstrates that substantial funding, precise modelling, and political awareness do not automatically translate into capacity to avert disaster. As ecological modelling matures, the critical challenge will be communicating business-useful information in formats that enable better decision-making and drives timely action.

It is crucial that climate change mitigations and adaptations in land management are compatible with reducing biodiversity decline. Climate Spheres’ own work extending climate model projection data into biophysical and agroclimatic variables will undoubtedly be enhanced by information and tools discussed at the event, and we look forward to collaborating with the ecological sector to support progress.



Authored by: Rosalind Haskins, PhD, Climate Spheres Research Director and Lead Climate Scientist

LinkedIn: linkedin.com/in/rosalind-haskins