Spatial Indicators of Post-Restoration Recovery in Wetland Ecosystems

Authors: Sophie Taddeo*, University of California - Berkeley, Iryna Dronova, University of California - Berkeley
Topics: Remote Sensing, Environment, Temporal GIS
Keywords: wetland, restoration, remote sensing, trajectory, object-based analysis
Session Type: Paper
Day: 4/12/2018
Start / End Time: 5:20 PM / 7:00 PM
Room: Bonaparte, Marriott, River Tower Elevators, 4th Floor
Presentation File: No File Uploaded


The scientific literature reports an important variability in the outcomes of wetland restoration, even among common habitat type or landscape context. A regional monitoring effort is critically needed to identify the causes of this variability and enhance the design of future projects. Yet, the consistent in-site monitoring of wetlands is made challenging by their remoteness and dynamic properties. Focusing on 20 restored wetlands and 5 reference sites of the Sacramento-San Joaquin Delta in California, US, we tested the use of public remote sensing data to characterize wetland trajectories, measure restoration progress, and identify constraints to wetland recovery. We conducted an object-based analysis on a dataset of aerial images captured by the National Agriculture Imagery Program spanning 11 years (2005-2016) to separate vegetated patches from open water and bare soil. Landscape metrics were then used to describe changes in the extent, distribution, greenness, and spectral heterogeneity of vegetated patches in both restored and reference sites. We grouped sites showing similar vegetation dynamics to assess the impact of landscape factors and site characteristics on wetland trajectories.
Preliminary results revealed the effect of initial site conditions and adjacent context on post-restoration development. Well-connected wetlands experienced a faster vegetation growth and wetlands restored during drought years showed greater annual variability in both greenness and spectral heterogeneity. Our study demonstrates that free remote sensing data can detect important patterns in wetland recovery. Studying these patterns enhances the current understanding of factors promoting wetland recovery and capacity to predict future restoration outcomes.


Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login