2.1 stress_analysis (if analysis_type includes “stress” or “comprehensive”)

Purpose: Identifies vegetation stress based on index thresholds

Structure:

result$analysis_results$stress_analysis

Contents for each index analyzed (e.g., NDVI, EVI, SIPI, GNDVI):

Threshold Structure:

result$analysis_results$stress_analysis$NDVI$thresholds_used
# $healthy: c(0.6, 1.0)         # NDVI 0.6-1.0 = healthy
# $moderate_stress: c(0.4, 0.6) # NDVI 0.4-0.6 = moderate stress
# $severe_stress: c(0.0, 0.4)   # NDVI 0.0-0.4 = severe stress

Interpretation:

• Healthy: Vegetation is growing normally, adequate water/nutrients
• Moderate Stress: Some stress present, may indicate water deficit or nutrient issues
• Severe Stress: Significant stress, requires immediate attention

Example Usage:

# Access stress results for NDVI
ndvi_stress <- result$analysis_results$stress_analysis$NDVI

# What percentage of my field is healthy?
cat(sprintf("Healthy vegetation: %.1f%%\n", ndvi_stress$healthy_percentage))

# What's the average NDVI?
cat(sprintf("Mean NDVI: %.3f\n", ndvi_stress$mean_value))

# Check all indices analyzed
names(result$analysis_results$stress_analysis)

2.2 growth_analysis (if analysis_type includes “growth” or “comprehensive”)

Purpose: Estimates crop growth stage and provides growth statistics

Structure:

result$analysis_results$growth_analysis

Contents:

# What growth stage is the crop in?
growth <- result$analysis_results$growth_analysis
cat(sprintf("Growth stage: %s (confidence: %.2f)\n", 
            growth$predicted_growth_stage,
            growth$stage_confidence))

# Get detailed NDVI statistics
ndvi_stats <- growth$NDVI
cat(sprintf("NDVI range: %.3f - %.3f\n", ndvi_stats$min, ndvi_stats$max))
cat(sprintf("NDVI variability (CV): %.3f\n", ndvi_stats$coefficient_of_variation))

2.3 yield_analysis (if analysis_type includes “yield” or “comprehensive”)

Purpose: Estimates yield potential using multiple vegetation indices

Structure:

result$analysis_results$yield_analysis

Contents:

Index Contributions Structure:

result$analysis_results$yield_analysis$index_contributions$NDVI
# $mean_normalized: 0.75    # Normalized contribution (0-1)
# $raw_mean: 0.68           # Raw mean NDVI value
# $raw_std: 0.12            # Raw standard deviation

Composite Yield Index Calculation:

1. Each index (NDVI, EVI, GNDVI, DVI, RVI) is normalized to 0-1 scale
2. Normalized values are averaged across all available indices
3. Result is a single 0-1 score where:
   • 0.0 = Very low yield potential
   • 0.5 = Medium yield potential
     
   • 1.0 = Maximum yield potential

Yield Potential Classifications:

For Corn:

• “Low”: composite_index < 0.3
• “Medium”: composite_index 0.3-0.6
• “High”: composite_index 0.6-0.8
• “Very High”: composite_index > 0.8

For Soybeans:

• “Low”: composite_index < 0.35
• “Medium”: composite_index 0.35-0.65
• “High”: composite_index 0.65-0.85
• “Very High”: composite_index > 0.85

For Wheat:

• “Low”: composite_index < 0.3
• “Medium”: composite_index 0.3-0.6
• “High”: composite_index 0.6-0.8
• “Very High”: composite_index > 0.8

Interpretation:

• Composite Yield Index (0.0-1.0): Higher values indicate better yield potential
• Yield Potential Class: Categorical assessment for easier interpretation
• Index Contributions: Shows which indices contributed and their individual scores
• Classification Confidence: Higher when composite_index is far from class boundaries (e.g., 0.2 or 0.9 are more confident than 0.5)

Example Usage:

# What's the yield potential?
yield <- result$analysis_results$yield_analysis
cat(sprintf("Yield Potential: %s (score: %.2f)\n", 
            yield$yield_potential_class,
            yield$composite_yield_index))

# Which indices contributed?
cat(sprintf("Based on %d indices: %s\n",
            yield$n_indices_used,
            paste(yield$indices_used, collapse = ", ")))

# Get individual index contributions
for (idx in names(yield$index_contributions)) {
  contrib <- yield$index_contributions[[idx]]
  cat(sprintf("%s: %.3f (raw: %.3f +/- %.3f)\n",
              idx, 
              contrib$mean_normalized,
              contrib$raw_mean,
              contrib$raw_std))
}

2.4 summary_statistics

Purpose: Basic statistical summary for all calculated indices

Structure:

result$analysis_results$summary_statistics

Contents for each index:

Plus Overall Summary:

result$analysis_results$summary_statistics$summary
# $total_indices_calculated: 6
# $indices_with_valid_data: c("NDVI", "EVI", "GNDVI", ...)
# $total_indices_requested: 6
# $success_rate: 100.0

Example Usage:

# Get statistics for all indices
stats <- result$analysis_results$summary_statistics

# NDVI statistics
ndvi_stats <- stats$NDVI
cat(sprintf("NDVI: %.3f +/- %.3f (range: %.3f to %.3f)\n",
            ndvi_stats$mean,
            ndvi_stats$std_dev,
            ndvi_stats$min,
            ndvi_stats$max))

# Check data quality
cat(sprintf("Coverage: %.1f%% (%d pixels)\n",
            ndvi_stats$coverage_percent,
            ndvi_stats$count))

2.5 validation (if reference_data provided)

Purpose: Validates analysis against ground truth data

Note: This component only appears if you provide reference_data parameter

Structure: TBD (depends on reference data format)

1. Identifying Problem Areas

# Find pixels with severe stress
ndvi <- result$vegetation_indices[["NDVI"]]
severe_stress <- ndvi < 0.4
plot(severe_stress, main = "Severe Stress Areas")

2. Comparing Multiple Fields

# Run analysis for multiple fields and compare
field1_result <- analyze_crop_vegetation(field1_data, crop_type = "corn")
field2_result <- analyze_crop_vegetation(field2_data, crop_type = "corn")

# Compare yield potential
cat(sprintf("Field 1 yield: %s (%.3f)\n", 
            field1_result$analysis_results$yield_analysis$yield_potential_class,
            field1_result$analysis_results$yield_analysis$composite_yield_index))
cat(sprintf("Field 2 yield: %s (%.3f)\n",
            field2_result$analysis_results$yield_analysis$yield_potential_class,
            field2_result$analysis_results$yield_analysis$composite_yield_index))

3. Time Series Analysis

# Analyze the same field at different dates
early_season <- analyze_crop_vegetation(june_data, growth_stage = "early")
mid_season <- analyze_crop_vegetation(july_data, growth_stage = "mid")
late_season <- analyze_crop_vegetation(august_data, growth_stage = "late")

# Track NDVI progression
ndvi_progression <- c(
  early_season$analysis_results$growth_analysis$NDVI$mean,
  mid_season$analysis_results$growth_analysis$NDVI$mean,
  late_season$analysis_results$growth_analysis$NDVI$mean
)
plot(1:3, ndvi_progression, type = "b",
     xlab = "Time Period", ylab = "Mean NDVI",
     main = "NDVI Progression Through Season")

Stress Detection

Thresholds are applied independently to each available index, with each index using biophysically appropriate breakpoints (Tucker, 1979; Hatfield et al., 2008):

Growth Stage Classification

Growth stage is predicted from mean NDVI using crop-specific thresholds. Statistics (mean, range, percentiles, CV) are additionally reported for EVI, GNDVI, and DVI. The threshold approach is conceptually consistent with Anyamba et al. (2021), who applied explicit per-crop NDVI cutoffs for large-scale MODIS-based crop monitoring, and with the MODIS + CDL crop monitoring framework of Akanbi et al. (2024).

Corn

Soybean

Wheat

These are literature-informed starting points. Exact NDVI boundaries vary by cultivar, sensor, region, and season. Calibration with local ground truth data is recommended.

Yield Potential (Composite Index)

Yield potential is assessed using a composite score that normalizes up to five indices (NDVI, EVI, GNDVI, DVI, RVI) to a 0–1 scale and averages them, reducing dependence on any single index. This multi-index approach is consistent with Bolton & Friedl (2013) and Mkhabela et al. (2011).

The composite yield index is a vegetation-based proxy, not a direct yield prediction. Correlation with actual harvested yield varies by crop, region, year, and management practice.

References

• Tucker, C.J. (1979). Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment, 8(2), 127–150. https://doi.org/10.1016/0034-4257(79)90013-0

• Hatfield, J.L., Gitelson, A.A., Schepers, J.S. & Walthall, C.L. (2008). Application of spectral remote sensing for agronomic decisions. Agronomy Journal, 100(S3), S-117–S-131. https://doi.org/10.2134/agronj2006.0370c

• Anyamba, A. et al. (2021). USA crop yield estimation with MODIS NDVI: Impact of the 2012 drought on corn production. Remote Sensing, 13(21), 4227. https://doi.org/10.3390/rs13214227

• Akanbi, O.D. et al. (2024). Integrating multiscale geospatial analysis for monitoring crop growth, nutrient distribution, and hydrological dynamics in large-scale agricultural systems. Journal of Geovisualization and Spatial Analysis, 8, 9. https://doi.org/10.1007/s41651-023-00164-y

• Bolton, D.K. & Friedl, M.A. (2013). Forecasting crop yield using remotely sensed vegetation indices and crop phenology metrics. Agricultural and Forest Meteorology, 173, 74–84. https://doi.org/10.1016/j.agrformet.2013.01.007

• Mkhabela, M.S. et al. (2011). Crop yield forecasting on the Canadian Prairies using MODIS NDVI data. Agricultural and Forest Meteorology, 151(3), 385–393. https://doi.org/10.1016/j.agrformet.2010.11.012