PEOPLE-ER Spectral Recovery

Authors: Sarah Zwiep, Melissa Birch, Marcos Kavlin

Reviewed/Edited by: Dr. Andy Dean, Dr. Nicholas Coops

Problem Statement:

Climate change effects are resulting in growing ecosystem degradation from increased disturbances such as wildfire and pest outbreaks. Ecosystem restoration (ER), a critical element of nature-based solutions, aims to return a degraded ecosystem to a stable state or reference condition. ER is important to reverse biodiversity loss, enhancing resilience to further disturbances. The need for ER on a large scale has been recognized by the United Nations sustainable development agenda and as part of the UN Decade on Ecosystem Restoration (2021-2030).

ER success is measured by ecosystem recovery, usually determined by three categories of ecosystem characteristics: vegetation structure, biodiversity, and ecological functioning (Ruiz-Jaen and Mitchell Aide 2005). Monitoring recovery is a challenge, as field-based measurements are impractical due to large spatial extents and a lack of resources and time. Field monitoring typically only occurs for short durations (<5 years), but studies have found resilience and recovery to occur on longer time scales (>10 years) (Pickell et al. 2016; Atkinson et al. 2022).

Remote sensing of ER offers a solution in continuous monitoring of large spatial and temporal extents. Free and open satellite observation programs such as Landsat and Sentinel-2 have increased ER monitoring potential (Wulder et al. 2012), with further development of spectral vegetation indices (VIs) and remotely sensed essential biodiversity variables (RS-EBVs). VIs and RS-EBVs enable the extraction of vegetation health and recovery metrics (Skidmore et al. 2021), subsequently allowing estimation of ecosystem vegetation structure, diversity, and functioning (Cabello et al. 2012; Cordell et al. 2017).

Trajectory-based time series analysis exploits temporal patterns in spectral values, proving to be effective for detecting and monitoring abrupt and gradual changes in ecosystem conditions (Banskota et al. 2014). However, previous analyses tend to explore only one spectral value or index, limiting the understanding of recovery dynamics, while change detection algorithms can be complex, unapproachable, and set with context-specific parameters, limiting more widespread applications (Cohen et al. 2018). This presents a barrier for effective use and interpretation of remotely sensed data for ER monitoring.

Objective:

The objective of the PEOPLE-ER Spectral Recovery tool is to provide an open source and multi-platform time series data analytics solution for both remote sensing research and ER monitoring purposes, which operates with freely available satellite imagery. It provides flexible methods for spectral recovery analysis, by providing users with the ability to select from a variety of spectral indices and recovery metrics as well as define reference or baseline conditions. The aim is to provide custom analyses ideal for widespread applications beyond site-specific contexts, with flexible reference conditions enabling direct integration with current ER initiatives and guidelines. The ability to calculate multiple per-pixel indices also provides remote sensing experts or researchers with the ability to easily produce spectrally derived products useful for further analysis.

PEOPLE-ER aims to provide the tool such that monitoring of ER is more accessible to a variety of users, enabling computation of indices and metrics by users that do not necessarily have a background in remote sensing or computer science. In providing a singular solution, spectral time series analysis becomes more approachable, increasing the opportunity to apply remote sensing techniques to ER monitoring.

Acknowledgements

This tool was developed within the Pioneer Earth Observation apPlications for the Environment Ecosystem Restoration (PEOPLE-ER) project financed by the European Space Agency (ESA). Members of the PEOPLE-ER consortium include: