syntropos 0.0.1

Computational engine for holistic spatiotemporal modeling and Geospatial Exploration Ecosystems.

Syntropos

Computational engine for holistic spatiotemporal modeling and Geospatial Exploration Ecosystems (EEGs).

---

I. Philosophy: The Living Territory

Traditional geographic information systems often treat the earth as a sterile, static grid—a mere receptacle for abiotic data. However, a territory is not a flat map; it is a highly organized, accumulating intersection of space, time, climate, and soil.

Syntropos (derived from Syntropy, the tendency towards organized complexity, and Topos, place) is the foundational software architecture designed to ingest chaotic, multi-terabyte streams of raw Earth observation data and distill them into highly structured, 4D spatiotemporal knowledge.

The engine provides the mathematical and computational rigor necessary to understand a territory holistically—modeling the macro-climatic teleconnections that drive local weather alongside the micro-topographic variations that dictate soil health, hydrology, and biological succession.

Core Pillars

• The Territory as a Dynamic Continuum: Syntropos rejects the isolation of spatial variables. It treats the territory as a teleconnected bioregion where every pixel interacts with its neighbors.
• Spatiotemporal Succession: Inspired by syntropic agriculture, the system treats time as the primary driver of complexity. It builds continuous, deeply indexed time-series data cubes to track the evolution of the landscape.
• Mathematical Rigor: Absolute spatial immutability. Through strict adherence to nested coordinate reference systems (like the COLGRID-9377 grid) and bitwise reduction algorithms, the engine preserves biological data integrity at the sub-pixel level.

---

II. Architecture: Decoupled by Design

Syntropos is engineered as a purely agnostic computational core, entirely decoupled from the geography it analyzes.

1. The Engine (syntropos): A universally deployable Python package. It handles STAC metadata parsing, lazy-loaded virtual raster reprojections, coordinate transformations, and statistical modeling. It is blind to specific regions, knowing only physics, math, and data structures.
2. The Node (e.g., OSESA): The regional implementation. A node (like the Observatorio de Suelos y Ecosistemas del Suroeste Antioqueño) provides the boundaries, the local configuration (config.yml), and the domain-specific research questions.

---

III. Technical Features

• Cloud-Optimized: Generates Cloud Optimized GeoTIFFs (COGs) with Zstandard (ZSTD) compression for high-performance I/O and web-native serving.
• Grid Alignment: Forced pixel alignment to regional nested grids, preventing fractional drift during reprojection.
• Intelligent Mosaicing: Implements bitwise OR operations for categorical quality masks (Fmask) to ensure the most conservative and accurate cloud/shadow detection in overlapping satellite swaths.
• Lazy Execution: Uses rasterio Virtual Raster Tables (VRTs) to minimize memory footprints when processing regional-scale daily mosaics.

---

IV. Installation

pip install syntropos