PyGeoFetch v2.6.2

PyGeoFetch version 2.6.2 Production Ready Multiple Provider Earth Observation and InSAR Platform

PyGeoFetch version 2.6.2 is a mature open source Python framework that unifies satellite data acquisition, predownload quality control, and advanced geospatial processing. It eliminates the fragmentation of multiple provider satellite APIs while introducing the first pure Python Windows native end to end InSAR processing chain. Designed for reproducibility and accessibility, it empowers researchers, government analysts, and students, particularly in the Global South, to execute complex Earth Observation workflows without relying on expensive commercial software or Linux high performance computing infrastructure.

The Preflight Gate serves as a novel predownload validation engine that evaluates stack viability before acquiring data. It automatically excludes scenes with insufficient area of interest coverage, validates Sentinel 1 TOPS burst synchronization to meet the strict five millisecond requirement, and implements orbit cascading fallbacks from precise to restituted to predicted orbits. This prevents the acquisition of terabytes of unusable data and saves significant bandwidth.

Windows Native InSAR Chain provides zero configuration SNAPHU integration that automatically detects operating system architecture and deploys precompiled binaries. This completely removes the notorious manual compilation barrier that has historically alienated Windows users from advanced InSAR processing.

Unified Multiple Provider API allows a single client search call to query over twenty providers, including Copernicus, USGS, Planetary Computer, AWS Earth, and NASA Earthdata, simultaneously. It returns a deduplicated and standardized SatelliteData object with automatic keyring based authentication.

Advanced InSAR and SAR Processing features a complete processing chain including orbit based geometric coregistration, Enhanced Spectral Diversity refinement, interferogram formation, topographic phase removal, and Goldstein filtering. The native SBAS time series inversion utilizes a Weighted Least Squares solver with native Phase Closure for unwrapping error correction and DEM Error Correction, which eliminates the need for external dependencies like MintPy. PS InSAR Hybrid Densification employs Amplitude Dispersion Index selection refined by temporal coherence, alongside NaN aware Atmospheric Phase Screen estimation. Large Deformation Monitoring utilizes amplitude based offset tracking with sub pixel parabolic refinement and signal to noise ratio based quality control for areas exceeding conventional phase limits, such as active mining sites and landslides. Atmospheric and Ionospheric Correction offers integrated support for ERA5, GACOS, and IONEX corrections.

Optical and Multi Sensor Processing includes a forty one step chainable pipeline for atmospheric correction, cloud masking, seventeen spectral indices, topographic correction, pan sharpening, and multi temporal compositing. Modern Geospatial Input and Output supports native Cloud Optimized GeoTIFF export, lazy loading via xarray and dask integration, and stateful provenance tracking.

Geospatial Analysis and Visualization enables interactive mapping via leafmap and MapViewer integration. It provides publication ready risk mapping with Bayesian uncertainty quantification and automated provenance manifests to ensure completely reproducible research.

Scientific Validation successfully detected Mexico City subsidence at negative thirty five point five centimeters per year, which is directionally consistent with the published benchmark of negative thirty nine point one centimeters per year reported by Cigna and Tapete in 2021, all while using a fraction of the data volume.

Operational Efficiency in benchmark testing reduced average data acquisition time by eighty seven percent and prevented the download of two point four terabytes of decorrelated or orbit deficient data via the Preflight Gate.
Production Scale has successfully processed over four hundred and fifty Sentinel 1 SLC pairs and more than one thousand two hundred optical scenes across diverse applications including mining subsidence in Obuasi Ghana, flood mapping, and urban expansion.

PyGeoFetch is platform independent for Python 3.9 and higher, and it is available via PyPI.
The core package for optical and multiple provider data access is installed via pip install pygeofetch.
Full InSAR capabilities, including native SBAS, PS InSAR, and offset tracking, are installed via pip install pygeofetch complete insar.

Advanced geospatial processing dependencies, such as rasterio and xarray, are installed via pip install pygeofetch geo.
Documentation and Support

This software is distributed under the MIT License, which is designed to maximize reuse, modification, and integration across both academic and operational contexts.

PyGeoFetch v2.6.2

PyGeoFetch v2.6.2: Production-Ready Multi-Provider Earth Observation & InSAR Platform

Executive Summary
PyGeoFetch v2.6.2 is a mature, open-source Python framework that unifies satellite data acquisition, pre-download quality control, and advanced geospatial processing. It eliminates the fragmentation of multi-provider satellite APIs while introducing the first pure-Python, Windows-native end-to-end InSAR processing chain. Designed for reproducibility and accessibility, it empowers researchers, government analysts, and students—particularly in the Global South—to execute complex Earth Observation (EO) workflows without relying on expensive commercial software or Linux/HPC infrastructure.

Core Innovations
The "Preflight Gate" (Bandwidth-Saving Quality Control): A novel pre-download validation engine that evaluates stack viability before acquiring data. It automatically excludes scenes with insufficient AOI coverage, validates Sentinel-1 TOPS burst synchronization (<5 ms requirement), and implements orbit-cascading fallbacks (Precise → Restituted → Predicted), preventing the acquisition of terabytes of unusable data.

Windows-Native InSAR Chain: Zero-configuration SNAPHU integration that automatically detects OS architecture and deploys pre-compiled binaries. This completely removes the notorious manual compilation barrier that has historically alienated Windows users from advanced InSAR processing.

Unified Multi-Provider API: A single client.search() call queries 22+ providers (Copernicus, USGS, Planetary Computer, AWS Earth, NASA Earthdata, etc.) simultaneously, returning a deduplicated, standardized SatelliteData object with automatic keyring-based authentication.

Comprehensive Processing Capabilities

Advanced InSAR & SAR Processing
Complete Processing Chain: Orbit-based geometric coregistration, Enhanced Spectral Diversity (ESD) refinement, interferogram formation, topographic phase removal, and Goldstein filtering.
Native SBAS Time-Series Inversion: Weighted Least Squares (WLS) solver with native Phase Closure (unwrapping error correction) and DEM Error Correction—eliminating the need for external dependencies like MintPy.
PS-InSAR Hybrid Densification: Amplitude Dispersion Index (ADI) selection refined by temporal coherence, with NaN-aware Atmospheric Phase Screen (APS) estimation.

Large-Deformation Monitoring: Amplitude-based offset tracking (speckle tracking) with sub-pixel parabolic refinement and SNR-based quality control for areas exceeding conventional phase limits (e.g., active mining, landslides).
Atmospheric & Ionospheric Correction: Integrated support for ERA5/PyAPS, GACOS, and IONEX/TEC corrections.
Optical & Multi-Sensor Processing

41-Step Chainable Pipeline: Atmospheric correction (DOS, 6S, Sen2Cor), cloud masking (SCL, Fmask), 17 spectral indices, topographic correction, pan-sharpening, and multi-temporal compositing.
Modern Geospatial I/O: Native Cloud Optimized GeoTIFF (COG) export, lazy loading via xarray/dask integration, and stateful provenance tracking.

Geospatial Analysis & Visualization
Interactive mapping via leafmap/MapViewer integration.
Publication-ready risk mapping with Bayesian uncertainty quantification.
Automated provenance manifests (provenance.yaml) ensuring 100% reproducible research.

Validated Research Impact
Scientific Validation: Successfully detected Mexico City subsidence at -35.5 cm/yr, directionally consistent with the published benchmark of -39.1 cm/yr (Cigna & Tapete, 2021, Remote Sensing of Environment), using a fraction of the data volume.
Operational Efficiency: In benchmark testing, reduced average data acquisition time by 87% and prevented the download of 2.4 TB of decorrelated or orbit-deficient data via the Preflight Gate.

Production Scale: Successfully processed 450+ Sentinel-1 SLC pairs and 1,200+ optical scenes across diverse applications including mining subsidence (Obuasi, Ghana), flood mapping, and urban expansion.

Installation
PyGeoFetch is platform-independent (Python 3.9+) and available via PyPI.

Core package (Optical + Multi-Provider Data Access)
pip install pygeofetch

Full InSAR capabilities (includes native SBAS, PS-InSAR, Offset Tracking)
pip install "pygeofetch[insar-full]"

Advanced geospatial processing dependencies (rasterio, xarray, etc.)
pip install "pygeofetch[geo]"

Documentation & Support
Documentation & Tutorials: https://appiahkubis14.github.io/pygeofetch-docs/
Source Code & Issue Tracker: https://github.com/EOCoreINT/pygeofetch
Benchmarking Scripts: Available in the benchmarks/ directory of the repository for independent verification of performance claims.

License
MIT License — Designed to maximize reuse, modification, and integration across both academic and operational contexts.

PyGeoFetch v2.6.2

PyGeoFetch version 2.6.2 Production Ready Multiple Provider Earth Observation and InSAR Platform

PyGeoFetch version 2.6.2 is a mature open source Python framework that unifies satellite data acquisition, predownload quality control, and advanced geospatial processing. It eliminates the fragmentation of multiple provider satellite APIs while introducing the first pure Python Windows native end to end InSAR processing chain. Designed for reproducibility and accessibility, it empowers researchers, government analysts, and students, particularly in the Global South, to execute complex Earth Observation workflows without relying on expensive commercial software or Linux high performance computing infrastructure.

The Preflight Gate serves as a novel predownload validation engine that evaluates stack viability before acquiring data. It automatically excludes scenes with insufficient area of interest coverage, validates Sentinel 1 TOPS burst synchronization to meet the strict five millisecond requirement, and implements orbit cascading fallbacks from precise to restituted to predicted orbits. This prevents the acquisition of terabytes of unusable data and saves significant bandwidth.

Windows Native InSAR Chain provides zero configuration SNAPHU integration that automatically detects operating system architecture and deploys precompiled binaries. This completely removes the notorious manual compilation barrier that has historically alienated Windows users from advanced InSAR processing.

Unified Multiple Provider API allows a single client search call to query over twenty providers, including Copernicus, USGS, Planetary Computer, AWS Earth, and NASA Earthdata, simultaneously. It returns a deduplicated and standardized SatelliteData object with automatic keyring based authentication.

Advanced InSAR and SAR Processing features a complete processing chain including orbit based geometric coregistration, Enhanced Spectral Diversity refinement, interferogram formation, topographic phase removal, and Goldstein filtering. The native SBAS time series inversion utilizes a Weighted Least Squares solver with native Phase Closure for unwrapping error correction and DEM Error Correction, which eliminates the need for external dependencies like MintPy. PS InSAR Hybrid Densification employs Amplitude Dispersion Index selection refined by temporal coherence, alongside NaN aware Atmospheric Phase Screen estimation. Large Deformation Monitoring utilizes amplitude based offset tracking with sub pixel parabolic refinement and signal to noise ratio based quality control for areas exceeding conventional phase limits, such as active mining sites and landslides. Atmospheric and Ionospheric Correction offers integrated support for ERA5, GACOS, and IONEX corrections.

Optical and Multi Sensor Processing includes a forty one step chainable pipeline for atmospheric correction, cloud masking, seventeen spectral indices, topographic correction, pan sharpening, and multi temporal compositing. Modern Geospatial Input and Output supports native Cloud Optimized GeoTIFF export, lazy loading via xarray and dask integration, and stateful provenance tracking.

Geospatial Analysis and Visualization enables interactive mapping via leafmap and MapViewer integration. It provides publication ready risk mapping with Bayesian uncertainty quantification and automated provenance manifests to ensure completely reproducible research.

Scientific Validation successfully detected Mexico City subsidence at negative thirty five point five centimeters per year, which is directionally consistent with the published benchmark of negative thirty nine point one centimeters per year reported by Cigna and Tapete in 2021, all while using a fraction of the data volume.

Operational Efficiency in benchmark testing reduced average data acquisition time by eighty seven percent and prevented the download of two point four terabytes of decorrelated or orbit deficient data via the Preflight Gate.
Production Scale has successfully processed over four hundred and fifty Sentinel 1 SLC pairs and more than one thousand two hundred optical scenes across diverse applications including mining subsidence in Obuasi Ghana, flood mapping, and urban expansion.

PyGeoFetch is platform independent for Python 3.9 and higher, and it is available via PyPI.
The core package for optical and multiple provider data access is installed via pip install pygeofetch.
Full InSAR capabilities, including native SBAS, PS InSAR, and offset tracking, are installed via pip install pygeofetch complete insar.

Advanced geospatial processing dependencies, such as rasterio and xarray, are installed via pip install pygeofetch geo.
Documentation and Support

This software is distributed under the MIT License, which is designed to maximize reuse, modification, and integration across both academic and operational contexts.