A comprehensive scientific tool for analyzing wheat phenology and estimating ground cover using NDVI observations integrated with real weather data from Open-Meteo API.
- Real Weather Data Integration: Uses Open-Meteo API (completely free, no API key required)
- Weather-Informed Growth Stages: Estimates growth stages using Growing Degree Days (GDD)
- Ground Cover Estimation: Calculates Fractional Vegetation Cover (FVC) and ground cover percentage
- Agricultural Stress Indices: Heat stress, cold stress, drought stress calculations
- Enhanced Visualizations: Multi-panel plots with weather data integration
- Location-Specific Analysis: Tailored for specific field locations
- Uncertainty Quantification: Realistic confidence intervals for predictions
pip install -r requirements.txtpython demo.pyfrom wheat_phenology_analyzer import WheatPhenologyAnalyzer
# Initialize analyzer
analyzer = WheatPhenologyAnalyzer(
ndvi_file='NDVI_ Treatment Parcel - 0-data-2025-07-04 15_11_14.csv',
sowing_date='03.10.2023',
harvest_date='30.07.2024',
geojson_file='field_location.geojson'
)
# Run analysis
analyzer.estimate_fvc_parameters(method='seasonal')
analyzer.interpolate_ndvi(method='balanced')
analyzer.estimate_growth_stages()
analyzer.create_visualization()
analyzer.save_results()
analyzer.generate_summary_report()wheat_phenology_weather_enhanced/
├── wheat_phenology_analyzer.py # Main analysis engine
├── demo.py # Demonstration script
├── setup.py # Installation script
├── requirements.txt # Python dependencies
├── field_location.geojson # Field coordinates (Bavaria, Germany)
├── NDVI_ Treatment Parcel - 0-data-2025-07-04 15_11_14.csv # NDVI observations
├── README.md # This documentation
├── wheat_phenology_results.csv # Output: Daily predictions
└── wheat_phenology_analysis.png # Output: Enhanced visualization
wheat_phenology_analyzer.py: Core analysis engine with weather integration, NDVI interpolation, and growth stage estimationfield_location.geojson: GeoJSON coordinates for Bavaria, Germany (49.2076°N, 10.6276°E)NDVI_ Treatment Parcel - 0-data-2025-07-04 15_11_14.csv: Sparse NDVI observations (5 data points)- Output files: Generated daily predictions and multi-panel visualizations
The analysis pipeline transforms sparse NDVI observations into comprehensive daily phenology data through seven sequential phases:
Input Data → Weather Integration → NDVI Processing → Growth Analysis → Visualization → Output
- Load GeoJSON coordinates for weather data fetching
- Fetch historical weather data from Open-Meteo API
- Load and validate sparse NDVI observations
- Calculate Growing Degree Days (GDD) with base temperature 0°C
- Literature-based: NDVI_soil = 0.15, NDVI_vegetation = 0.85
- Data-driven: Based on observed NDVI min/max values
- Seasonal: Using early and mid-season patterns
- Apply cubic spline interpolation for smooth daily time series
- Calculate realistic confidence intervals (15% uncertainty factor)
- Ensure physiologically valid NDVI range [0,1]
- Compute FVC and ground cover percentages
- Calculate cumulative GDD from sowing date
- Apply wheat-specific GDD thresholds:
- Sowing (0), Emergence (50), Tillering (200), Stem Elongation (500)
- Booting (800), Heading (1000), Flowering (1200), Grain Filling (1400)
- Maturity (1800), Harvest (2000)
- Identify exact transition dates and handle overlapping stages
- Heat Stress: Maximum temperature > 30°C
- Cold Stress: Minimum temperature < -5°C
- Drought Stress: Consecutive days without precipitation
- Correlate stress periods with growth stages
- Generate multi-panel plots with growth stage markers
- Include NDVI trends, ground cover, weather conditions, and GDD
- Add vertical lines and labels for each growth stage transition
- Apply professional formatting and high-resolution output
- Combine all parameters into comprehensive daily dataset
- Generate summary statistics by growth stage
- Create detailed CSV output and visualization files
The pipeline transforms sparse NDVI observations (5-20 data points) into comprehensive daily phenology data through weather-enhanced analysis and robust uncertainty quantification.
- Real weather data integration via Open-Meteo API
- Growing Degree Days calculation with actual temperature data
- Location-specific analysis and stress assessment
- Weather-informed growth stage timing
- Physiologically constrained confidence intervals
- Realistic uncertainty estimates based on observed data variability
- Avoids unrealistic extrapolation beyond valid NDVI ranges [0,1]
- Handles sparse data scenarios appropriately
- Multiple FVC parameter estimation methods
- Configurable growth stage thresholds
- Adaptable interpolation approaches
- Comprehensive stress index calculations
- Input: GeoJSON coordinates → Weather API → NDVI CSV
- Processing: GDD calculation → NDVI interpolation → Growth stages
- Output: Daily predictions → Visualizations → Summary reports
- API fallback mechanisms for connection failures
- Input validation and format checking
- Robust interpolation for edge cases
- Quality control for output data
- API: Open-Meteo Archive API
- URL: https://archive-api.open-meteo.com/v1/archive
- Cost: Completely free
- Registration: Not required
- API Key: Not required
- Temperature (mean, min, max)
- Precipitation
- Relative humidity
- Atmospheric pressure
- Wind speed
- Cloud cover
- Growing Degree Days (GDD)
- Agricultural stress indices
- Field: Bavaria, Germany
- Coordinates: 49.2076°N, 10.6276°E
- Analysis Period: October 3, 2023 to July 30, 2024
- Balanced Interpolation: Combines physiological knowledge with smooth transitions
- Confidence Intervals: Realistic uncertainty quantification based on data variability
- Physiological Constraints: Ensures NDVI values stay within valid range [0, 1]
- Fractional Vegetation Cover: Calculated from NDVI using linear mixing model
- Ground Cover Percentage: FVC expressed as percentage
- Parameter Estimation: Literature-based, data-driven, or seasonal methods
- Weather-Informed: Uses Growing Degree Days (GDD) thresholds
- Stages: Sowing, Emergence, Tillering, Stem Elongation, Booting, Heading, Flowering, Grain Filling, Maturity, Harvest
- GDD Thresholds: Each stage has specific GDD requirements
- Heat Stress: Temperature > 30°C
- Cold Stress: Temperature < -5°C
- Drought Stress: Consecutive days without precipitation
- Optimal Conditions: Temperature 10-25°C with precipitation
Daily predictions including:
- NDVI interpolated values with confidence intervals
- FVC and ground cover percentage
- Growth stage assignments
- Weather data (temperature, precipitation, humidity, etc.)
- Growing Degree Days
- Agricultural stress indices
Multi-panel plot showing:
- NDVI time series with confidence intervals and growth stage markers
- Ground cover percentage with uncertainty bounds
- Weather conditions (temperature and precipitation)
- Growing Degree Days with stage threshold lines
The generated visualization includes:
- Panel 1 (Top Left): NDVI time series with confidence intervals (shaded area) and vertical growth stage markers
- Panel 2 (Top Right): Ground cover percentage showing vegetation coverage over time
- Panel 3 (Bottom Left): Weather conditions with temperature (line) and precipitation (bars)
- Panel 4 (Bottom Right): Cumulative Growing Degree Days with horizontal threshold lines for each growth stage
Each growth stage transition is marked with:
- Vertical dashed lines indicating exact transition dates
- Color-coded labels showing stage names
- Precise timing for each growth phase
The visualization provides a comprehensive view of wheat development throughout the growing season, integrating NDVI observations, weather conditions, and phenological stages.
The pipeline generates a high-resolution multi-panel visualization showing the complete wheat phenology analysis:
Note: The actual visualization file (wheat_phenology_analysis.png) is generated when running the analysis and contains the complete multi-panel plot with growth stage markers, confidence intervals, and weather data integration.
growth_stages = {
'Sowing': {'gdd_threshold': 0},
'Emergence': {'gdd_threshold': 50},
'Tillering': {'gdd_threshold': 200},
'Stem Elongation': {'gdd_threshold': 500},
'Booting': {'gdd_threshold': 800},
'Heading': {'gdd_threshold': 1000},
'Flowering': {'gdd_threshold': 1200},
'Grain Filling': {'gdd_threshold': 1400},
'Maturity': {'gdd_threshold': 1800},
'Harvest': {'gdd_threshold': 2000}
}- Literature: NDVI_soil = 0.15, NDVI_vegetation = 0.85
- Data-driven: Based on observed NDVI min/max values
- Seasonal: Based on early and mid-season NDVI patterns
- Real Data: Actual weather data, not synthetic
- Location-Specific: Tailored to field location
- Weather-Informed: Growth stages based on accumulated heat units
- Stress Awareness: Identifies weather-related stress conditions
- Physiological Knowledge: Incorporates wheat growth patterns
- Uncertainty Quantification: Realistic confidence intervals for predictions
- Multiple Methods: Various interpolation and estimation approaches
- Robust Estimation: Designed for sparse data scenarios
- Crop Management: Timing of agricultural operations
- Yield Prediction: Growth stage-based yield estimation
- Stress Monitoring: Early detection of weather stress
- Research: Phenology studies and climate impact assessment
- Formula: NDVI = (NIR - Red) / (NIR + Red)
- Range: -1 to +1 (typically 0 to 1 for vegetation)
- Interpretation: Higher values indicate more vegetation
- Formula: FVC = (NDVI - NDVI_soil) / (NDVI_vegetation - NDVI_soil)
- Range: 0 to 1
- Interpretation: Proportion of ground covered by vegetation
- Formula: GDD = max(0, (T_max + T_min) / 2 - T_base)
- Base Temperature: 0°C for wheat
- Interpretation: Accumulated heat units for plant development
pip install -r requirements.txtpython setup.pyfrom wheat_phenology_analyzer import WheatPhenologyAnalyzer
analyzer = WheatPhenologyAnalyzer(
ndvi_file='your_ndvi_data.csv',
sowing_date='03.10.2023',
harvest_date='30.07.2024',
geojson_file='field_location.geojson'
)
analyzer.estimate_fvc_parameters(method='seasonal')
analyzer.interpolate_ndvi(method='balanced')
analyzer.estimate_growth_stages()
analyzer.create_visualization()
analyzer.save_results()# Custom FVC parameters
analyzer.estimate_fvc_parameters(method='literature')
# Different interpolation method
analyzer.interpolate_ndvi(method='linear')
# Custom output filename
analyzer.save_results('custom_results.csv')
analyzer.create_visualization('custom_plot.png')CSV file with columns:
phenomenonTime: Date/time of observation (YYYY-MM-DD HH:MM:SS)NDVI: NDVI values (0-1 range)
GeoJSON file with polygon coordinates for the field location.
- Sowing date: DD.MM.YYYY (e.g., '03.10.2023')
- Harvest date: DD.MM.YYYY (e.g., '30.07.2024')
-
API Connection Errors: The tool automatically falls back to synthetic weather data if the API is unavailable.
-
Missing Files: Ensure all required files are present in the working directory.
-
Date Format Errors: Use the correct date format (DD.MM.YYYY) for sowing and harvest dates.
-
NDVI Data Issues: Verify NDVI values are within the valid range [0, 1].
- The tool is optimized for sparse NDVI data (5-20 observations)
- Weather data fetching may take 10-30 seconds depending on internet connection
- Analysis typically completes in under 1 minute for a full growing season
This tool is designed for agricultural research and crop management applications. Contributions are welcome for:
- Additional weather data sources
- Enhanced interpolation methods
- New agricultural stress indices
- Improved visualization options
- Additional crop types beyond wheat
- Open-Meteo API for providing free weather data
- Agricultural research community for phenology models
- Remote sensing community for NDVI applications
For questions or support, please refer to the code documentation or create an issue in the repository.