Skip to main content

Humboldt Ocean Watch

Python
Streamlit
ocean monitoring
SST
Experimental thermal-monitoring application for Niño 1+2.

Humboldt Ocean Watch is an experimental Python and Streamlit system for reproducible daily sea-surface-temperature diagnostics in the Niño 1+2 region.

Repository

Current scientific scope

The implemented operational scope is thermal. The system works with cached SST and climatological products to calculate and validate metrics such as:

  • daily SST and anomalies;
  • standardized anomalies;
  • daily and seven-day changes;
  • persistence and threshold-area diagnostics;
  • spatial and temporal feature summaries;
  • thermal-event detection;
  • connected thermal patches, tracks, and event families;
  • quality-control and provenance outputs;
  • optional scientific briefs grounded in locally calculated facts.

Important limitation

This software does not provide an official Coastal El Niño classification, causal attribution, forecast, or biological, fishery, economic, or social impact assessment. It does not replace official oceanographic products.

Language models, when enabled, interpret validated local outputs; they do not calculate the scientific metrics.

Architecture

The application separates:

  1. data loading and geographic validation;
  2. climatology and anomaly calculations;
  3. deterministic scientific analytics;
  4. quality control and provenance;
  5. cached products for the interface;
  6. optional grounded natural-language interpretation.

The dashboard reads prepared products rather than silently downloading or recalculating data during normal interaction.

Development status

The application is suitable for methodological demonstration and continued scientific development. It remains labelled experimental until data-refresh infrastructure, broader-domain climatologies, operational governance, and external validation are mature enough for a stronger claim.