AUTHOR=Wang Zhanpeng , Ke Yinghai , Lu Dan , Zhuo Zhaojun , Zhou Qingqing , Han Yue , Sun Peiyu , Gong Zhaoning , Zhou Demin 

TITLE=Estimating fractional cover of saltmarsh vegetation species in coastal wetlands in the Yellow River Delta, China using ensemble learning model

JOURNAL=Frontiers in Marine Science

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2022.1077907

DOI=10.3389/fmars.2022.1077907

ISSN=2296-7745

ABSTRACT=Saltmarshes in coastal wetlands provide important ecological services, while they have been suffering from area loss and degradation in ecological functions. Previous research has utilized satellite remote sensing imagery for the mapping and classification of saltmarsh vegetation. However, the “hard classification” might cause serious “mixed pixel” problems as the saltmarsh landscapes are characterized by spatial heterogeneities at fine scales. Sub-pixel fractional cover estimation of saltmarsh vegetation at species level are required to better understand the distribution and canopy structure of saltmarsh vegetation. In this study, we presented an approach framework for estimating and mapping the fractional cover of major saltmarsh species in the Yellow River Delta, China based on time series Landsat 8 Operational Land Imager data. To solve the problem that the coastal area is frequently covered by clouds, we adopted the recently developed virtual image-based cloud removal (VICR) algorithm to reconstruct missing image values under the cloud/cloud shadows over the time series Landsat imagery. Then, we developed an ensemble learning model, which incorporates Random Forest Regression (RFR), K-Nearest Neighbor Regression (KNNR) and Gradient Boosted Regression Tree (GBRT) based on temporal-spectral features derived from the time-series cloudless images to estimate the fractional cover of major vegetation types, i.e., Phragmites australis, Suaeda salsa and the invasive species, Spartina alterniflora. High spatial resolution imagery acquired by the Unmanned Aerial Vehicle and Gaofen-6 satellites were used for reference sample collections. The results showed that our approach successfully estimated the fractional cover of each saltmarsh species, with the average R-squared value of 0.891 and the root mean square error below 11%. Through four scenarios of experiments, we found that the ensemble learning model is advantageous over each individual model, with the R-square increased by 4.94% compared to KNNR model and 2.33% compared to GBRT model. When the images during key months were absent, cloud removal for the Landsat images considerably improved the estimation accuracies. The fractional cover maps reveal the relationships between the invasion period and the cover of S. alterniflora.