// JAN 2025 — FEB 2025
Data Analyst
UQ Gas & Energy Transition Centre
Brisbane, Australia
PythonTime-series analysisMachine LearningDeep LearningSensor dataStatistical Analysis
What I worked on
A short-term analyst role focused on measurement reliability for the centre's methane-emission research. Methane sensors drift with temperature, humidity, and wind — and that drift matters when the data is used for sustainability reporting.
The pipeline
- Built a time-series pipeline to quantify methane-measurement accuracy drift against temperature, humidity, and windspeed
- Extracted emission patterns, trends, and anomalies across the centre's monitoring data
- Validated 2D/3D anemometer measurements through sensor agreement analysis and error/variance decomposition
- Refactored and reviewed parts of the existing codebase to improve reliability, reproducibility, and analysis performance
Why it mattered
Drift quantification turns "the sensor said X" into "the sensor said X, with this confidence interval given the conditions" — which is what makes the resulting research defensible.
Some details under NDA.