The land-sea interface mapping: China's coastal land covers at 10 meters for 2020
Topics: China
, Remote Sensing
, Land Use and Land Cover Change
Keywords: Coast, Remote sensing, Land cover, Machine learning, Google Earth Engine, China
Session Type: Virtual Guided Poster Abstract
Day: Sunday
Session Start / End Time: 2/27/2022 09:40 AM (Eastern Time (US & Canada)) - 2/27/2022 11:00 AM (Eastern Time (US & Canada))
Room: Virtual 1
Authors:
Miao Li, Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies,Tsinghua University, Beijing 100084, China
Bin Chen, Division of Landscape Architecture, Faculty of Architecture, The University of Hong Kong, Hong Kong SAR, China
Bing Xu, Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies,Tsinghua University, Beijing 100084, China
,
,
,
,
,
,
,
Abstract
The land-sea interface (LSI) generally referred to as coastal areas, is among the most productive ecosystems. In China, the rapid urbanization along the coastal regions has dramatically reshaped the geomorphology, which arouse a series of environmental concerns. Multi-source and multi-scale remote sensing has enabled large-scale monitoring of the complicated and highly dynamic LSI over time. However, the inconsistent data quality, mapping standard, modeling method, and spatiotemporal coverage of these datasets have often yielded different results, making it challenging for an accurate and comprehensive socio-ecological assessment. Addressing this challenge, we report a new high-resolution LSI mapping: China’s Coastal Land Covers (CCLC) at 10 meters for the year 2020. We first establish an LSI classification system containing 12 essential coastal land cover types that outline anthropogenic influences and biophysical processes in China. We then build a coastal land cover sample library across the LSI of China using a semi-empirical generation approach based on existing datasets and visual interpretations. We further develop region-specific random forest models using features extracted from time-series of Sentinel-1 and Sentinel-2 imagery, topography, nighttime light, and population data. Based on the initial pixel-based classification results, we refine them using the integration of object-based and knowledge-based post-processing procedures. The validation results indicated that the CCLC achieved an overall accuracy of 83%, outperforming that of ESA WorldCover (69%), Esri Land Cover (56%), GlobeLand30 (57%), and GLC_FCS30 (53%) on China’s coast. This CCLC map can provide comprehensive information on uncovering coastal human-natural interactions and support integrated coastal conservation and management.
The land-sea interface mapping: China's coastal land covers at 10 meters for 2020
Category
Virtual Guided Poster Abstract
Description
This abstract is part of a session. Click here to view the session.
| Slides