I am an environmental data analyst applying hydrological, geospatial and statistical methods to flood, drought and climate-risk questions in the UK.
- Hydrology and rainfall-runoff modelling
- Flood-frequency analysis and catchment diagnostics
- Drought, low-flow and water-resources assessment
- GIS, spatial analysis and environmental mapping
- Climate-change risk, GHG emissions, carbon and Net Zero pathways
- Statistical analysis, uncertainty assessment and environmental data visualisation
- Machine learning for prediction, classification and model interpretation
An investigation of observed high flows in an urban Scottish catchment using sub-hourly discharge, gridded rainfall and open geospatial data. The study compares annual-maximum and peaks-over-threshold frequency estimates and examines flood seasonality, rainfall-runoff response, terrain and regional high-flow behaviour. Across the fitted statistical distributions, the estimated 100-year flood ranges from 73.9 to 102.2 m³/s, demonstrating the importance of model-selection uncertainty when extrapolating beyond the observed record.
An observed-data assessment of drought duration and accumulated flow deficit at Fakenham, supported by regional precipitation and nearby Chalk-groundwater records. The 1991-2020 flow-duration curve gives Q90 = 0.274 m³/s. Mean annual days below Q90 increase from 12.8 in 1967-1985 to 30.4 in 2005-2023, but the block-bootstrap interval for the early-to-late difference includes zero. The study therefore characterises changing low-flow exposure without attributing a climate-driven trend where abstraction and sampling uncertainty remain relevant.
An analysis of final DESNZ territorial emissions statistics and the reduction pace implied by the UK's 2030 and 2035 Nationally Determined Contributions. UK territorial emissions fell by 417.4 MtCO2e, or 52.8%, between 1990 and 2024, with electricity supply and industry accounting for 66.0% of the net reduction. From the final 2024 NDC-scope estimate, the corresponding compound reduction rates are 6.25% per year to 2030 and 7.93% per year to 2035. The analysis keeps historical observations, policy targets, diagnostic pathways and reporting scope uncertainty separate.
A daily demand forecasting study combining Yorkshire Water meter observations with Met Office HadUK-Grid regional weather. Chronological evaluation compares recent-demand baselines with Ridge regression across internal and external meters in two anonymised areas. Weather produces small, inconsistent gains; the largest main-comparison improvement is a 3.36% reduction in MAE for Area 2 external-meter median consumption, using same-day observed weather. The study examines extreme readings and changing meter coverage, and distinguishes retrospective results from operational forecasting evidence.
Hydrology: flood-frequency analysis, low-flow analysis, rainfall-runoff
modelling and hydroclimatic diagnostics
Spatial analysis: GIS, GeoPandas, Rasterio, Shapely and catchment analysis
Data science: Python, R, pandas, NumPy, SciPy, xarray and scikit-learn
Scientific practice: uncertainty assessment, data provenance, testing, Git
and GitHub Actions
My work combines observed environmental data, clearly stated assumptions, traceable data provenance and explicit uncertainty assessment. I distinguish exploratory evidence from conclusions suitable for operational or design decisions.
Based in the UK and interested in opportunities across flood risk, water resources, environmental analytics, climate resilience, GIS and applied data science.