# Recurrence Analysis of Vegetation Time Series and Phase Transitions in   Mediterranean Rangelands

**Authors:** Teodoro Semeraro, Norbert Marwan, Bruce K. Jones, Roberta Aretano,, Maria Rita Pasimeni, Irene Petrosillo, Christian Mulder, and Giovanni Zurlini

arXiv: 1705.04813 · 2025-02-13

## TL;DR

This study uses recurrence analysis on vegetation time series to understand how fire disturbances affect the dynamics and predictability of Mediterranean rangelands, aiding better ecosystem management.

## Contribution

It introduces recurrence analysis with sliding windows to analyze vegetation dynamics and phase transitions in fire-affected Mediterranean rangelands.

## Key findings

- Fire simplifies vegetation structures and dynamics, making them more regular and predictable.
- Grasslands are more predictable than forests post-fire due to different successional responses.
- Recurrence analysis helps map landscape transitions and supports ecosystem management strategies.

## Abstract

Mediterranean rangelands should be conceived as socio-ecological landscapes (SEL) because of the close interaction and coevolution between socio-economic and natural systems. A significant threat to these Mediterranean rangelands is related to uncontrolled fires that can cause potential damages due to the reduction or even the loss of ecosystems. Our results show that time series of forest and grassland for unburned and burned areas are characterized by both periodic and chaotic components. The fire event caused a clear simplification of vegetation structures as well as of SEL dynamics that is more regular and predictable after the burning and less chaotic. However grassland evolution could be more predictable than forest considering the effect of fire disturbance on successional cycles and stages of the two land-cover types. In particular, we applied recurrence analysis with sliding temporal windows three-year length on the original time series. This analysis indicates that grasslands and forests behaved similarly in correspondence with the burning, although their phase states slowly diverge after fire. Recurrence is useful to study the vegetation recovery as it enables mapping landscape transitions derived from remote sensing. The approach helps stakeholders to draw landscape interventions and improve management strategies to sustain the delivery of ecosystem services.

---
Source: https://tomesphere.com/paper/1705.04813