🌤️ Forecast Model Verification (Alpha)

Welcome to my weather model verification project!

This repository contains R scripts designed to compare numerical weather prediction models (currently ECMWF and ICON-GR) against actual meteorological observations (OBS) across Greek stations.

🚧 Disclaimer: Early Alpha Version! Please note that this project is in its very early stages. I am actively learning, building, and refactoring as I go. The code you see here works, but it is not the final, polished version. Expect messy loops, frequent updates, and structural changes!

---

What does this do?

The goal of this project is to evaluate the performance of weather forecasts by calculating standard statistical metrics for lead times from 0h to 48h.

Currently, the scripts calculate RMSE (Root Mean Square Error) and ME (Mean Error) for the following meteorological parameters:

• 🌡️ TEMP (Temperature)
• 💧 TD (Dew Point)
• 🌬️ WS (Wind Speed)
• ☁️ TCC (Total Cloud Cover)
• 📉 MSLP (Mean Sea Level Pressure)

Repository Structure

To keep things organized, the project expects a specific folder structure. We use the here package in R, so everything is beautifully portable!

• scripts/ : Contains all the .R scripts. This is where the magic happens.
• data/ : This is where you should place your raw .txt input files (Note: The actual data files are ignored via .gitignore to keep the repo lightweight).
• output/ : Automatically generated by the scripts.
  • output/docs/ : Saves the merged data frames and calculated metrics as .txt files.
  • output/plots/ : Saves the final comparison plots as .png images.

Future Plans (To-Do)

As I continue to work on this, I plan to:

• Optimize the code (reduce repetition and use functions).
• Code and outputs translation to English

---

Feel free to explore the code, and remember: it's a work in progress!