WIP: Structural Dynamics of UK Biodiversity Records (2000–2025)

Conducted an exploratory structural analysis of UK biodiversity occurrence records spanning 2000–2025. The project investigates concentration dynamics across datasets using multiple inequality metrics (Top 1%, Top 10%, Gini coefficient, Herfindahl–Hirschman Index), identifying long-run dominance patterns and structural regime shifts.

Implemented log-based growth decomposition to separate total record growth into extensive (number of active datasets) and intensive (mean occurrences per dataset) margins. Results show that pre-2020 growth was primarily driven by intensity expansion among dominant datasets, with evidence of organic decentralisation beginning prior to the 2020 structural break caused by dataset inactivity.

Additional analysis included tier-based seasonality comparison, robustness checks excluding inactive datasets, and structural break identification. Findings demonstrate how aggregation effects and dataset composition materially influence perceived ecological trends.

Project implemented in Python using pandas, NumPy, and matplotlib. Emphasis placed on reproducible workflows, multi-metric validation, and robustness testing.

• Source Code
• View Notebook