wind-forecast

application for wind forecast data analysis

There are a variety of wind forecast models such as ECMWF or GFS, which often differ on the forecast for a given location. How do you determine which model is the most historically accurate at a given location? How do you predict which model will be the most accurate forecast / which you should rely on for your next free flight?

This application uses the free OpenMeteo API to obtain wind forecast data from 2 weather forecast models (ECMWF and GFS), and two types of sensors to obtain measured wind data: Tempest and Holfuy sensors. Note that the OpenMeteo GFS model is enriched with HRRR model data for forecasts < 3 days from the requested date. The application codebase can be found on Skip McGee's Wind-Forecast repository.

install the requirements:

1) follow the os-specific install instructions at: https://pypi.org/project/mysqlclient/ 2) python3 -m venv venv 3) source venv/bin/activate or venv\Scripts\activate on Windows. 4) pip install --upgrade pip setuptools wheel 5) pip install -r requirements.txt

(required) create your database connection file:

open database/.env

add the following customized variables:

340DBHOST='mydbhost'
340DBUSER='myusername'
340DBPW='mypw'
340DB='mydb'

(optional) add your sensor api keys:

open app/.env

add the following customized variables:

HOLFUY_STATION='1151'
HOLFUY_TOKEN='mytoken'
TEMPEST_STATION='134520'
TEMPEST_TOKEN='mytoken'
TEMPEST_STATION_NAME='Sandia Soaring Peak Tempest'

make sure your network or local db is set up:

1) confirm database name, user, pass 2) source DDL.sql to create tables and data

To actually run the app:

1) source venv/bin/activate or if on Windows venv\Scripts\activate 2) python3 app.py

Syntax to run via Gunicorn:

gunicorn --config gunicorn.py wsgi:app

skipmcgee / wind-forecast

readme