Work

Aeris Ops: AI-Powered Smart Weather Intelligence, Disaster Prediction and Climate Analytics Platform

Year: 2026
Type: article
Abstract: Weather intelligence becomes operationally useful only when forecasts, historical context, hazard signals, uncertainty and recommended actions are connected without overstating what the data can prove. This updated review paper examines AERIS OPS as an implemented web prototype rather than as a validated disaster-forecasting service. The reviewed project combines a React/TypeScript client, Express/tRPC services, Drizzle persistence contracts, Open-Meteo forecast and air-quality adapters, ERA5-based historical analytics, deterministic multi-hazard screening, explainability and risk dashboards, Leaflet location intelligence, GDACS/NASA live hazard-layer adapters, agriculture guidance, alerts, incident review and PDF/CSV export. A key design feature is explicit provenance: outputs are labelled LIVE, FALLBACK, HISTORICAL, SAMPLE, RULE-BASED or UNAVAILABLE, while analytical triggers are kept separate from official warnings. The project snapshot passed TypeScript checking, production build, and 45 automated tests; these results demonstrate software-level verification, not predictive accuracy or real-world impact. The paper adds an implementation-specific research-gap matrix covering the integration-to-action gap, provenance and uncertainty gap, hazard data-availability gap, explainability evaluation gap, safe notification gap and prototype-to-evidence gap. Future work should establish labelled regional event datasets, calibrated models, exposure and vulnerability layers, human-centred evaluations, official operational integrations and prospective field validation. Index Terms— weather intelligence, explainable AI, impact-based forecasting, disaster risk management, climate analytics, data provenance, rule-based screening, uncertainty communication, human-centred decision support.
Language: English
Open Access status: hybrid
OpenAlex ID: W7234869218
FWCI:
Cites:
Cited by:
Related to: