🌍📡 Planet API – Advanced Agronomic & Environmental Intelligence Workflows

This repository contains production-grade Jupyter notebooks demonstrating how to acquire, process, validate, analyze, and interpret Planet satellite data using Python.

The workflows implemented here are designed for real-world agronomic intelligence, environmental monitoring, ESG analytics, and geospatial data science.

This is not just API usage — it is an end-to-end analytical framework for remote sensing applications.

---

📂 Repository Structure

🛰️ api-planet-imagery.ipynb

Complete workflow for PlanetScope Surface Reflectance (8-band) imagery.

---

🔐 Data Acquisition & Ordering

• Secure authentication

• AOI definition (GeoJSON or Shapefile → GeoJSON conversion)

• Search by:

  • Temporal range
  • PSScene
  • AnalyticMS_SR_8b

• Order creation with:

  • AOI clipping
  • Surface Reflectance products
  • Automated download

---

🧠 Understanding Planet Assets

Example:

20250515_133533_10_2506_3B_AnalyticMS_SR_8b_clip.tif

Breakdown:

• 3B → Level 3B (orthorectified)
• AnalyticMS_SR → Surface Reflectance
• 8b → 8 spectral bands
• udm2 → pixel usability mask

Bands confirmed programmatically:

coastal_blue
blue
green_i
green
yellow
red
rededge
nir

---

🛡️ Data Quality & Masking (UDM2)

New functionality includes:

• Automatic detection of matching UDM2 mask
• Valid pixel mask generation
• Cloud/shadow/no-data removal
• Safe masking before index calculation

This ensures agronomically reliable outputs.

---

🌱 Spectral Index Processing (Modular Functions)

process_ndvi(tif_path)

• Uses Red & NIR bands
• Applies valid mask
• Returns masked NDVI array

process_ndre(tif_path)

• Uses RedEdge & NIR
• Sensitive to nitrogen & canopy structure
• Mask-aware implementation

---

🎨 Visualization & Distribution Analysis

show_ndvi(path)

show_ndre(path)

Each function:

• Displays index map with agronomically meaningful palette

  • NDVI → brown → yellow → green
  • NDRE → yellow → orange → dark green

• Generates Seaborn distribution plot (histogram + KDE)

• Uses clipping only for visualization, not for computation

Also explains:

• Why density ranges vary
• Why index values are clipped for display
• Interpretation of pixel distributions

---

🚜 Agronomic Applications (Imagery Notebook)

• Crop vigor mapping
• Early stress detection
• Irrigation performance assessment
• Field heterogeneity detection
• Support for variable-rate application
• Input for ML models

---

🌱 api-planet-planetary-variables.ipynb

Advanced workflow for Planetary Variables (L4) — derived products optimized for time-series monitoring.

---

📡 Product Discovery & Quota Inspection

New additions:

• Query available products via:

GET /my/subscriptions/products

• Inspect:

  • resource_ids
  • quota_total
  • quota_used
  • quota_units
  • supports_reservation

Helps prevent accidental overconsumption.

---

🧾 Subscription-Based Time Series

Instead of downloading rasters:

• Create subscription
• Consume /results
• Parse returned statistics
• Build Pandas DataFrame

Robust handling added for:

statistics = None

Prevents runtime failures.

---

📊 Supported Planetary Variables

💧 Soil Water Content (SWC)

• Unit: volumetric fraction (m³/m³)
• Example: 0.13 → 13% soil volumetric water content

---

🌡️ Land Surface Temperature (LST)

• Unit: Kelvin

• Example:

  • 288.82 K → 15.67 °C
  • 301.40 K → 28.25 °C

Converted interpretation implemented.

---

🌾 Biomass Proxy

• Dimensionless proxy index (0–1 scale typical)

• Example:

  • 0.523 → moderate-to-high canopy biomass

---

📈 DataFrame Construction Pipeline

Converts raw results:

{
  "statistics": [{"name":"mean","value":0.274}]
}

Into structured dataset:

datetime_utc	mean_swc	valid_percent	source_id

Sorted, typed, ready for:

• Trend analysis
• Anomaly detection
• Dashboarding
• Predictive modeling

---

🔬 Agronomic Interpretation Layer

The notebooks now include interpretation logic for:

• SWC in irrigated systems
• LST thermal stress thresholds
• Biomass proxy vigor inference
• NDVI vs NDRE canopy stage differentiation
• Early nitrogen stress detection

Designed specifically for crops such as:

• Tobacco (irrigated)
• Soybean
• Corn
• Perennial crops
• ESG-monitored areas

---

🚀 Why This Repository Is Different

This project integrates:

• API engineering
• Raster processing
• Pixel masking
• Spectral analysis
• Time series modeling
• Agronomic interpretation
• Cost-awareness (quota monitoring)

It is a remote sensing intelligence stack, not just a demo.

---

⚠️ Important Notes

• Valid Planet API credentials required
• Product access depends on contract
• Large AOIs may be restricted
• Always monitor quota before large batch operations

---

🎯 Target Audience

• Precision agriculture professionals
• Environmental monitoring specialists
• ESG analytics teams
• Remote sensing engineers
• Data scientists working with geospatial data
• Consultants building agronomic intelligence systems

---

🌍 Final Perspective

These notebooks demonstrate how to transform:

Satellite data → Clean raster → Valid pixels → Spectral indices → Structured time series → Agronomic insight → Decision support.

This is the bridge between remote sensing and actionable intelligence.