Skip to main content

Computational Science

Machine learning tools for automated analysis of environmental time-series data. Published openly for use by any institution globally.
Planned

The analysis layer across every dataset

Every HICS instrument produces time series — and turning raw readings into insight takes computation. This programme builds open machine-learning and analysis tools for environmental data: cleaning and gap-filling sensor records, detecting anomalies, and modelling the patterns in air, ground, and sky. Every tool is published openly, doubling as teaching material.

open
tools & notebooks
published in Python, reusable by anyone
all data
one analysis layer
air quality, seismic, muon, thermal
Time-series ML

Models for environmental time series — forecasting, gap-filling, and separating signal from sensor noise.

Anomaly detection

Flagging unusual readings automatically — feeding the data quality flags on the open dataset.

Open notebooks

Analyses shipped as Jupyter notebooks built around real HICS data — also the backbone of the scientific computing courses.

Current status

Planned as a cross-cutting layer. It grows with the datasets — see the open data API and the scientific computing courses under Education.