This tool predicts the optimal harvest time for alfalfa fields by analyzing yield and quality based on multispectral Harmonized Landsat and Sentinel-2 (HLS) and SAR (Sentinel-1) satellite imagery. An economic model recommends the best harvest time, considering the predicted yield, quality, weather forecast data, and user-defined parameters. This web platform, developed by the GDSL group at Purdue University, efficiently delivers crop information and serves as a reference for informed harvest decisions.
- Frontend: React
- Backend: FastAPI
- Database: SQLite3
This project is a full-stack web application designed to facilitate complex data interactions between a frontend built with React, a backend API developed with FastAPI, and a SQLite3 database. The application integrates with external services such as Google Earth Engine (GEE), NASA's Earthdata, and Google Maps to provide advanced visualization and data processing capabilities.
-
Frontend:
- Developed with React for a responsive, dynamic user interface.
-
Backend:
- Built using FastAPI, delivering a robust, high-performance API.
-
Database:
- SQLite3, offering lightweight and efficient data storage.
-
External Integrations:
- Google Earth Engine (GEE): For downloading and processing Sentinel-1 data.
- NASA Earthdata: Accessing Harmonized Landsat and Sentinel-2 (HLS) satellite data.
- Google Maps API: Providing map visualization.
- Install dependencies:
pip install -r backend/requirements.txt
- Run the backend server:
uvicorn backend.main:app --reload
- Install dependencies:
npm install
- Start the development server:
npm start
Configure the following environment variables:
GOOGLE_MAPS_API_KEY=your_google_maps_api_key
EARTHDATA_API_TOKEN=your_nasa_earthdata_token
GEE_CREDENTIALS=path_to_your_gee_credentials_file