turbulence-fetcher

Computes mechanical, thermal, shear, and convective turbulence forecasts for UK lat/lon locations.

Updated Aug 6, 2026
One-click install
npx skills add https://github.com/vrai/ga_weather_briefing_agentic --skill turbulence-fetcher-vrai
Or copy as Structured Prompt for Agentβ–Ό
Please help me install this Agent Skill.
Skill: turbulence-fetcher
Source: https://github.com/vrai/ga_weather_briefing_agentic/tree/main/skills/turbulence-fetcher
Command: npx skills add https://github.com/vrai/ga_weather_briefing_agentic --skill turbulence-fetcher-vrai

SYSTEM DOCUMENTATION & REQUIREMENTS

πŸ’‘ This Skill includes scripts (resource) components.

What problem does it solve? Pilots planning UK general aviation flights need turbulence detail beyond what METAR/TAF gusts or the F215 chart provide, and assembling model data from multiple sources by hand is slow and error-prone. ## Core Features & Use Cases - Four turbulence mechanisms: Classifies mechanical (wind/gust), thermal (convective velocity scale w*), shear (including low-level jet and inversion signatures), and convective (CAPE-led) turbulence into light/moderate/strong buckets across 3-hour daylight periods. - Multi-source data fusion: Combines Open-Meteo ICON and GFS model fields with the optional Convek soaring API, falling back to a Deardorff w* estimate labelled "(estimated)" when no Convek key is configured. - Pre-digested output: Returns a formatted TURBULENCE ANALYSIS text block with period tables and headline summaries, ready to quote directly into a GA weather briefing. - Use Case: Given a route midpoint like "EGKA:50.835602:-0.297222", produce a same-day turbulence forecast showing the thermic window, boundary-layer top, morning shear risk, and CAPE peaks for a briefing. ## Quick Start Ask the agent to run the turbulence forecast for a location such as EGSL at 51.895, 0.448 and include the resulting analysis in today's flight briefing.

Frequently Asked Questions about turbulence-fetcher

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get a turbulence forecast for a UK airfield?β–Ό

Run the script with a LABEL:LAT:LON argument, for example: uv run scripts/get_turbulence.py --location "EGSL:51.895:0.448". It returns JSON containing a pre-formatted TURBULENCE ANALYSIS text block covering mechanical, thermal, shear, and convective turbulence.

What data sources does the turbulence forecast use?β–Ό

It fetches winds, temperatures, CAPE, and radiation from the Open-Meteo ICON model, boundary layer height and lifted index from Open-Meteo GFS, and optionally thermal strength and convergence from the Convek soaring API. None of the Open-Meteo calls require an API key.

Do I need a Convek API key to use this skill?β–Ό

No, the Convek API key is optional. Without CONVEK_API_KEY set in the skill's .env file, thermal strength is estimated from ICON radiation and GFS boundary-layer depth via a Deardorff w* calculation, and those values are labelled "(estimated)" in the output.

Is this turbulence forecast certified for flight planning?β–Ό

No. The light/moderate/strong classification thresholds are v1 rules of thumb not calibrated against observed ride-quality reports. Treat the output as model-derived heuristics, never as a replacement for PIREPs or F215 turbulence areas.

Why does the turbulence output say "not computed" for some mechanisms?β–Ό

A mechanism shows "not computed" when its required input data was unavailable, such as missing boundary-layer or CAPE fields from the model fetches. The design degrades gracefully per mechanism rather than failing the whole analysis.

How do I run the turbulence.py unit tests?β–Ό

From the skill's scripts directory, run python3 test_turbulence.py or python3 -m unittest test_turbulence -v. The 26 tests use only the standard library since turbulence.py is a pure computation module with no I/O or network access.