Satellite Monitoring
in development
The Satellite Monitoring Module is Envorum's high-tech analytical complex designed for the long-term verification of reforestation and soil regeneration projects. Unlike one-time reports, this module creates a Digital Twin of a land plot and tracks its condition over a 3-year (36-month) period using the Sentinel-2 satellite constellation and Google Earth Engine (GEE) computing power.
1. Initiation Criteria
The monitoring module is activated automatically when a Planting event is created, provided the following conditions are met:
Scale: The target number of trees is $\ge$ 100 units.
Compliance: A valid planting permit (Permit) is uploaded to the event’s document section.
Geometry: The organizer defines the perimeter by marking all corner coordinates (Polygon). This ensures maximum precision during satellite imagery "clipping" in GEE, excluding noise from neighboring territories.
2. Engineering Analysis Layers
The system minimizes errors and external noise through a multi-stage data processing pipeline:
📊 Reference Area Control (Differential Analysis)
The satellite analyzes not only the target polygon but also a "reference area" within a 100–200 meter radius. This allows the system to isolate seasonal factors such as a general regional greening after rains and identify the pure impact of the planted trees.
⛈️ Climate & Weather Integration
Integration with meteorological grids (CHIRPS and ERA5) enriches reports with precipitation and temperature anomaly data. If saplings grow slowly due to drought, the AI arbiter classifies this as a climate risk rather than organizer negligence.
🛡️ Quality Filtering (Cloud & Shadow Masking)
The S2cloudless algorithm automatically discards images with cloud cover $> 30%$, shadows, or snow. If a high-quality image cannot be obtained, the monitoring iteration is postponed until the next clear satellite pass.
🧬 Adaptive Spectral Monitoring
To ensure maximum precision across the lifecycle, the system dynamically switches primary spectral indices:
Year 1 (Survival): Focus on SAVI (Soil Adjusted Vegetation Index) to eliminate soil background interference and verify initial rooting.
Year 2 (Growth): Comparison of NDVI vs SAVI to distinguish tree growth from inter-row weed interference.
Year 3 (Health): Shift to NDRE or EVI to assess chlorophyll levels and canopy density as the foliage thickens.
3. Three-Year Cycle and Biological Standards
The Gemini algorithm interprets incoming data based on established forest life cycles and scientific growth expectations:
0–12 mo
SAVI (Soil Correction)
~82%
Rooting period. Focus on stabilizing indices despite small canopy size and transplant stress.
12–24 mo
NDVI + SAVI (Growth)
~50–70%
Formation of primary foliage. Distinguishing trees from weeds via inter-index comparison.
24–36 mo
NDRE / EVI (Health)
~40–60%
Stabilization phase. Assessing chlorophyll quality and final ecosystem sustainability.
4. Autonomous Decentralized Verification (AD-Veri)
To maintain 100% autonomy when data confidence is low (Confidence Score < 0.6) due to weather or contradictory indices, the platform triggers the AD-Veri protocol:
Auto-Tasking: The system automatically generates a paid bounty for the nearest Scouts (within a 20km radius).
Ground Truth: The Scout performs on-ground photo/video verification anchored to the polygon's corner coordinates.
Synthesis: Ground-level data is merged with satellite data to form an indisputable proof of the planting status.
5. Deliverables (Output)
Every 3, 6, 9, 12, 18, 24, 30, and 36 months, the system generates:
Executive Summary: A human-readable report generated by Gemini regarding the forest's health status.
JSON Metadata: A structured data array for smart contracts and external auditors.
Impact Certificate: A final digital verification issued after the 36-month mark, confirming the established ecosystem.
Last updated