Projects & Tools

Overview

This page showcases computational tools, workflows, and research projects I’ve developed for InSAR analysis, geospatial data processing, and Earth observation applications.


InSAR Processing Workflows

HPC-Scale InSAR Time-Series Processing Pipeline

Technologies: Python, ISCE, MintPy, SLURM, Dask

Developed a scalable workflow for processing multi-terabyte SAR datasets on high-performance computing clusters:

  • Automated ISCE processing for Sentinel-1 and ALOS-2 data stacks
  • Parallel processing using SLURM job arrays for efficient compute resource utilization
  • Integration with MintPy/MiaplPy for advanced time-series analysis
  • Covariance-aware phase linking for signal enhancement in challenging terrain
  • Automated quality control and validation pipelines

Applications: Landslide monitoring in the Andes, subsidence studies, tectonic deformation analysis


Tropospheric Correction Framework

Technologies: Python, ERA5, MERRA-2, PyAPS, NumPy

Implemented multiple atmospheric correction strategies for InSAR:

  • Statistical approaches using empirical phase-elevation relationships
  • Physics-based ray-tracing with global weather models (ERA5, MERRA-2)
  • Comparative analysis of correction methods for different climate zones
  • Automated integration into time-series workflows

Impact: Significantly improved deformation signal quality in high-relief mountainous terrain


Geospatial Data Science Tools

InSAR Deformation Classification Pipeline

Technologies: Python, scikit-learn, GeoPandas, xarray

Developed machine learning workflows for automated deformation pattern recognition:

  • Spatial clustering of InSAR time-series velocities
  • K-means and DBSCAN-based deformation regime classification
  • Automated landslide boundary detection and mapping
  • GIS-ready output generation for ArcGIS Pro and QGIS

Data Quality Control & Validation Tools

Technologies: Python, pandas, pytest, CLI development

Built automated pipelines for geospatial data validation:

  • Schema and content consistency checks
  • Encoding normalization and format validation
  • Fuzzy matching using TF-IDF and text similarity methods
  • Structured logging and reproducible run reporting
  • CLI tools for indexing and reliability scoring

Developed at: Statista Strategy (2025)


Multi-Sensor Integration Projects

Multi-Temporal LiDAR Differencing

Technologies: LAStools, Python, GDAL, CloudCompare

Implemented workflows for terrain change detection:

  • LiDAR point cloud processing and DEM generation
  • Multi-temporal DEM differencing for displacement quantification
  • Integration with InSAR results for cross-validation
  • Automated change detection and volume calculations

GNSS + InSAR Integration

Technologies: GAMIT/GLOBK, Python, NumPy

Developed workflows for combining geodetic datasets:

  • GNSS time-series processing and quality control
  • Co-registration of GNSS and InSAR measurements
  • Joint inversion for earthquake deformation modeling
  • Uncertainty propagation and error analysis

Research Software Contributions

InSAR Processing Optimization

Contributed to improved processing efficiency for community tools:

  • Optimized StaMPS and TRAIN workflows for parallel computing
  • Developed Python wrappers for automated batch processing
  • Created documentation and tutorials for reproducible analysis

Optical Offset Tracking

Technologies: COSI-Corr, Python, GDAL

Implemented fast displacement mapping for rapid deformation events:

  • Automated offset tracking for landslide and glacier studies
  • Integration with optical satellite data (Sentinel-2, Landsat)
  • Multi-scale analysis for different deformation rates

Specialized Applications

Groundwater Depletion Monitoring

Project: Minab Coastal Plain, Iran
Technologies: InSAR, GRACE data, aquifer modeling, Python

Integrated satellite observations with hydrological modeling:

  • InSAR-derived subsidence time series
  • GRACE-based groundwater storage change analysis
  • Aquifer parameter estimation through inverse modeling
  • Spatiotemporal correlation analysis

Status: Published at ICCC Workshop 2025


Earthquake Sequence Analysis

Project: 2022 Zagros Mountains, Iran
Technologies: InSAR, Pyrocko, Kite, Grond, ObsPy

Contributed to multi-sensor geodetic analysis:

  • Co-seismic deformation mapping with Sentinel-1
  • Fault parameter inversion using InSAR observations
  • Integration with seismological data
  • Published in Seismica (2023)

Landslide Kinematics Characterization

Project: Central Andes, Argentina
Technologies: InSAR time-series, spatial statistics, Python

Comprehensive study of slow-moving landslides:

  • Multi-temporal deformation analysis (2014-2024)
  • Seasonal pattern recognition and hydrological correlation
  • 2D/3D decomposition of displacement fields
  • Landslide classification based on kinematics

Status: Ph.D. dissertation research (ongoing)


Open Source & Community Contributions

GitHub Projects

Visit my GitHub profile for code repositories and tutorials:

  • InSAR processing scripts and workflows
  • Python utilities for geospatial analysis
  • Tutorial notebooks for remote sensing applications
  • Documentation and best practices

Knowledge Sharing

Blog Posts: Technical tutorials on InSAR processing, atmospheric corrections, and Python workflows
Presentations: Conference talks and workshop materials on InSAR analysis
Mentoring: Supporting students and early-career researchers in remote sensing


Technical Stack

Primary Tools

  • InSAR Processing: ISCE, MintPy, MiaplPy, Dolphin, StaMPS, SARscape
  • Programming: Python, MATLAB, R, Bash
  • Data Science: NumPy, pandas, GeoPandas, scikit-learn, xarray, Dask
  • GIS: ArcGIS Pro, QGIS, PostgreSQL/PostGIS
  • Version Control: Git, GitHub
  • HPC: SLURM, parallel computing workflows

Development Practices

  • Reproducible research workflows
  • Test-driven development (pytest)
  • CLI tool development
  • Documentation and code commenting
  • Continuous integration for data pipelines

Future Projects

I’m currently exploring:

  • Deep Learning for InSAR: Applying CNNs for automated deformation detection and phase unwrapping
  • Cloud-Based Processing: Migrating workflows to Google Earth Engine and AWS for scalability
  • Real-Time Monitoring: Developing near-real-time landslide warning systems
  • Multi-Temporal Fusion: Combining SAR, optical, and LiDAR for comprehensive change detection

Collaboration Opportunities

I’m interested in collaborating on:

  • InSAR applications for natural hazard monitoring
  • Machine learning for Earth observation
  • Geospatial data science projects
  • Open-source tool development

📧 Get in touch: mohseniaref@uni-potsdam.de