Extracting spatiotemporal characteristics and succession pattern of
native and invasive plant species using plant phenophase and
multi-source remote sensing images in the Yellow River delta, China
Abstract
The native and invasive species in the Yellow River Delta were examined
for their spatiotemporal characteristics and succession pattern. First,
the appropriate Sentinel-2 and Landsat-8 images from 2018 were selected
according to phenological characteristics. A random forest algorithm was
used to verify the image spectral band significance and separability
using selected images to determine the native and invasive species.
Then, the spatiotemporal variation of habitat structure of native and
invasive species is discussed in depth from the perspective of landscape
ecology. Finally, the expansion direction and expansion mode of S.
alterniflora were further analyzed, and main results were obtained as
follows. (1) At the medium-high resolution multi-spectral image level,
the accuracy of different vegetation community extractions can be
improved by taking into consideration both the vegetation phenology and
the spectral features of remote sensing images. (2) Sentinel-2 images
with red edge bands have obvious advantages in vegetation community
extraction as compared to Landsat-8 images (Sentinel-2, OA=82.86%,
Kappa coefficient=0.79; Landsat-8, OA=78.77%, Kappa coefficient=0.74).
(3) The expansion pattern of the S. alterniflora community became
spatially continuous, more regularized and aggregated overtime. (4) The
expansion in the north shore mainly faces to the sea, and the south bank
mainly faces to the land, and this phenomena is closely related to the
sedimentation of the Yellow River Delta. Marginal and external expansion
both occurred, but marginal expansion predominated. The results from
this study have important theoretical and scientific value for the
environmental protection and sustainable development of the entire
Yellow River Delta.