Document Type
Dissertation
Degree
Doctor of Philosophy (PhD)
Major/Program
Earth Systems Science
First Advisor's Name
Shimon Wdowinski
First Advisor's Committee Title
Committee chair
Second Advisor's Name
Daniel Gann
Second Advisor's Committee Title
Committee member
Third Advisor's Name
Dean Whitman
Third Advisor's Committee Title
Committee member
Fourth Advisor's Name
Assefa Melesse
Fourth Advisor's Committee Title
Committee member
Fifth Advisor's Name
Malek Adjouadi
Fifth Advisor's Committee Title
Committee member
Keywords
Synthetic Aperture Radar, remote sensing, wetlands, hydrology, water depth, vegetation, polarization, optical remote sensing, Landsat, Sentinel-1, machine learning, mangrove, canopy height, forest, coastal area, Hurricane Irma, random forest, neural network, airborne LiDAR, G-LiHT, EDEN, sawgrass
Date of Defense
6-28-2022
Abstract
The Everglades wetlands located in the South Florida is characterized by large geographical extent, heterogeneous and diverse plant community, and provides various ecosystem services. Hydrology is the most important abiotic factor controlling wetland functions, whereas monitoring hydrology for such a large extent is a challenging task. Water gauges, which is able to provide accurate and high temporal frequency (hourly) water level measurements, are main tool for hydrologic monitoring. However, the point measurements only provided limited hydrologic information in space. Though the measurements can be interpolated to produce water surfaces, the accuracy of the map depends on the density of gauge network. The first two studies in this dissertation investigate the possibility to monitor wetland hydrology using Synthetic Aperture Radar (SAR ) observations. The first study comprehensively investigated Sentinel-1 C-band Synthetic Aperture Radar (SAR) observations’ sensitivity in response to water depths variations. This study mainly found that SAR backscatter observations are linearly correlated with changes of water depths in wetlands, for which the type of correlation varies with vegetation types. Sparse woody vegetation is characterized by positive backscatter-water depth linear correlation; medium dense herbaceous by a combination of positive-negative correlation; sparse herbaceous by a negative correlation. The second study detects the significant water-depth increase (SWDI) based on the linear relationships found in the first study. The study develops a new algorithm using pre-, during-, and post-hurricane backscatter observations to classify the areas with SWDI, which is defined by water depth increase more than 12 cm. The algorithm successfully classifies the vast extent of SWDI during the landing of Hurricane Irma on September 12, 2017, and the decrease of SWDI areas for the following two months. The third study used the Sentinel-1 SAR data, combined with the other two types of remote-sensing data: optical and airborne laser scanning (ALS) data, to map the canopy height of mangroves in the Everglades National Park before and after Hurricane Irma with machine/deep learning models. We used the traditional random forest and developed an innovative hybrid neural network model to map mangrove canopy heights. Both models achieved high accuracy with Mean Absolute Error (MAE) of ~2m. We estimated the canopy loss using pre- and post-hurricane canopy height maps, and found the taller mangroves are the most impacted ones by the hurricane. The novel knowledge and algorithms developed in the three studies could be widely used in global wetland for hydrological and ecological monitoring.
Identifier
FIDC010739
Previously Published In
Chapter 3 has been published in
Zhang, B., Wdowinski, S., Gann, D., Hong, S., Sah J.P., Spatiotemporal variations of wetland backscatter: The role of water depth and vegetation characteristics in Sentinel-1 dual-polarization SAR observations. Remote Sensing of Environment. 2022, 270, 112864.
The publisher of the journal paper is Elsevier
Chapter 4 has been published in
Zhang, B., Wdowinski, S. and Gann, D., 2022. Space-Based Detection of Significant Water-Depth Increase Induced by Hurricane Irma in the Everglades Wetlands Using Sentinel-1 SAR Backscatter Observations. Remote Sensing, 14(6), p.1415.
The publisher of the journal paper is MDPI
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Zhang, Boya, "Hydrological and Ecological Synthetic Aperture Radar (SAR) Applications for the Everglades Wetland Ecosystem" (2022). FIU Electronic Theses and Dissertations. 5059.
https://digitalcommons.fiu.edu/etd/5059
Included in
Computational Engineering Commons, Computer Engineering Commons, Terrestrial and Aquatic Ecology Commons
Rights Statement
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/
This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).