Weather-I-Go

Have you ever had your plans ruined by bad weather? Weather-I-Go uses NASA data with AI to give you friendly weather insights based on your ‘perfect weather’ preferences!

Comment

Weather-I-Go 🌦️

Inspiration

Most weather apps only give numbers like temperature, humidity, or wind speed. We wanted to build something that explains weather in plain, friendly language and tells you if it matches your personal idea of a good day. Using NASA Earth data, long-term climate projections, and AI, we thought it would be cool to connect science with everyday life.

What it does

Weather-I-Go takes your location, date, and “perfect weather” preferences (like sunny and 25°C), and then:

  1. Pulls real weather data from NASA’s POWER API and MateoMedics Weather API, which lets us access weather projections up to the year 2100.
  2. Uses AI to explain the weather in plain text, so it’s easy to understand.
  3. Compares your perfect weather to the actual or projected conditions, so you know if the day matches what you’re hoping for.

How we built it

Tools & tech we used: Frontend: React Native (with Expo), TypeScript, AsyncStorage, React Native Calendars, multi-slider. Backend: Node.js with Express, Firebase for user events, APIs (NASA POWER, Meteomatics, OpenStreetMap). AI: Google Gemini for natural, friendly weather explanations.

Challenges we ran into

  • NASA POWER limitations: The API only provides historical and near real-time data, not forecasts. This made it challenging to predict future dates, which was one of our main goals.
  • Data cleanup: The dataset often returned placeholder values like -999, which required careful handling and validation logic.
  • Google Earth Engine integration: Learning how to query and process geospatial datasets in GEE while connecting it with our backend was a steep learning curve.
  • Frontend + Backend integration: Making React (frontend) and Node/Express (backend) communicate smoothly took more debugging time than expected.
  • Version control issues: As a team, we had to resolve multiple Git conflicts and learn how to collaborate effectively on the same codebase.
  • AI pipeline: Getting the backend and AI response pipeline to work smoothly.

Accomplishments that we're proud of

  • Successfully built a full-stack app that connects NASA POWER data with an AI explanation system.
  • Cleaned and structured messy API outputs into a user-friendly format.
  • Integrated React frontend + Node.js backend so users can interact with their weather preferences in real time.
  • Learned to collaborate as a team, even when running into Git conflicts.

What we learned

  • How to use APIs effectively, including their limitations (like NASA POWER not having future forecasts).
  • The difference between historical data vs. forecast data, and why that matters when designing apps.
  • How to integrate a frontend and backend into one working pipeline instead of treating them as separate projects.
  • Hands-on experience with React, Tailwind, Node.js, Express, and Google Earth Engine, and how they fit together.

What's next for Weather-I-Go

  • Add more interactivity, like maps where users can click a location and instantly see AI weather insights.
  • Explore NASA’s GEOS FP or other forecast datasets to support predictions for days to months ahead.
  • Enhance the AI explanations so they’re not just text, but also visual (charts, comparisons, alerts).
  • Turn the prototype into a polished mobile app for everyday use.

Built With

Share this project:

Updates