This page highlights (1) open-source tools and teaching tutorials developed, (2) my publications:

Open-Source Software and tutorials

BFAST — Breaks For Additive Seasonal and Trend

BFAST is an open-source method for detecting and characterising change in time series data — originally developed for satellite image time series, and now widely used in ecology, forestry, and land monitoring.

What it does:

  • Detects abrupt changes (breakpoints) in trend and seasonal components
  • Works with dense, multi-year time series (e.g. MODIS, Landsat, Sentinel)
  • Handles missing data and noise typical in remote sensing products

Where to find it:

Current maintainer of the package: Dr. Dainius Masiliūnas.

Geo-Scripting — Open Course Materials

An open-source course developed at Wageningen University teaching geospatial scripting in R, Python, and Bash within a Linux environment. Winner of the Excellent Teaching Award at Wageningen University in 2017 and 2018.

What you will learn:

  • Geospatial data handling with R and Python
  • Raster and vector analysis at scale
  • Scripting workflows in Linux and Bash
  • Working with satellite data (Landsat, Sentinel, MODIS)

Access the materials:

The materials are fully open-source and continue to be used internationally.

Publications and a few key papers

For a complete publication list:

A few key papers:

  • Ecosystem resilience monitoring using Earth Observation (2024) Bathiany S. et al. (co-author) — Surveys in Geophysics (2024) DOI: 10.1007/s10712-024-09833-z A collaborative synthesis from the International Space Science Institute (ISSI) Workshop on Tipping Points. Addresses challenges and opportunities in using EO data for resilience monitoring — directly connected to the TipMIP intercomparison project.

  • Detecting change in satellite time series — BFAST (2010) Verbesselt J., Hyndman R., Newnham G. — Remote Sensing of Environment (2010) The foundational paper behind the BFAST method. Introduces a framework for decomposing satellite time series into trend, seasonal, and remainder components — and detecting structural breaks within each.

  • Remotely sensed resilience of tropical forests (2016) Verbesselt J. et al. — Nature Climate Change (2016) DOI: 10.1038/nclimate3108 Demonstrates that satellite time series can be used to detect early warning signals of forest resilience loss — before visible degradation occurs. One of the highest-impact papers from my group, with implications for global forest monitoring and tipping point research.