Things I've built

Active

hazard-alerter

A daily weather-hazard forecast, triggered from your phone.

• A GitHub Actions workflow (daily_forecast.yml) that runs on demand via workflow_dispatch.
• Triggered by an Apple Shortcut that grabs the phone's current location, extracts the ZIP code, and posts the dispatch with that ZIP as an input.

github.com/higborn01/hazard-alerter

In progress

boring-log-app

An open-source field app for geotechnical boring logs.

• Handles soil boring logs, bedrock/rock core logs, and well logs, with built-in soil classification aligned to EPA / state DEP / USGS standards.
• Ships the full symbology library rather than a curated subset, reusing existing agency symbol assets as-is.
• Starting with NJDEP's jurisdiction-specific symbols (well-construction, environmental), with other states to follow.
• Offline-first for field data collection, with an office mode; a Rust/C core via FFI handles performance-critical pieces like USCS classification and DIGGS/AGS parsing, with Python on the server side for heavy validation and reporting.
• Data model built first: a Postgres schema shaped to map onto the DIGGS standard.

Complete

coral-sea-dee2020

Reproducing a coral-based superposed epoch analysis from a 2020 Science paper on volcanic eruptions and ENSO.

• Three parallel implementations — the original MATLAB reference, plus Python and R ports — of the Superposed Epoch Analysis (SEA) from Dee et al. 2020.
• Tests whether El Niño-like responses follow major tropical volcanic eruptions, using the Palmyra coral δ¹⁸O record.
• Cross-validated across all three languages to 4 decimal places, byte-verified against the paper's official supplementary materials.
• Finding: the composite stays inside the 95% confidence band at every post-eruption lag — no statistically significant response detected.

github.com/higborn01/coral-sea-dee2020