摘要
为揭示近三十年江苏沿海地区变化特征,以1988—2018年江苏沿海湿地变化为研究对象,利用11期Landsat卫星影像,分别采用最大似然分类法和随机森林分类法对研究区土地覆盖信息进行提取并分析湿地变化特征。结果表明,基于最大似然分类得出的总体精度和Kappa系数分别为77.96%和0.67,基于随机森林分类得出的总体精度和Kappa系数分别为80.77%和0.69,随机森林分类精度更高;近30年江苏沿海湿地面积共减少498.46 km2,耕地和建筑用地呈增长状态,城镇人口变化与湿地面积变化有较强的负相关性,表明湿地开垦、城市化进程等人类活动是驱使江苏沿海地区湿地面积不断缩减的主要因素。研究结果可为江苏沿海湿地变化分析、湿地环境保护提供参考。
In order to reveal the characteristics of coastal area change in Jiangsu province in the past 30 years, the land cover information of the wetland in the study area was extracted by using the 11 Landsat satellite images from 1988 to 2018. The results show that the total accuracy and kappa coefficient based on maximum likelihood classification are 77.96% and 0.67 respectively. The total accuracy and kappa coefficient based on random forest classification are 80.77% and 0.69 respectively, which indicates that the accuracy of random forest classification is higher. In addition, the coastal wetland area of Jiangsu province has decreased 498.46 km~2 in recent 30 years, and cultivated land and construction land have been growing. The urban population change has a strong negative correlation with wetland area change, which indicates that human activities such as wetland reclamation and urbanization process are the main factors driving the wetland area to continue to reduce in Jiangsu coastal areas. The results can provide theoretical and data support for wetland change along the coast of Jiangsu province.
作者
俞峂柠
郭炯甫
王泽乾
魏浩翰
YU Tong-ning;GUO Jiong-fu;WANG Ze-qian;WEI Hao-han(College of Civil Engineering,Nanjing Forestry University,Nanjing Jiangsu 210037,China)
出处
《现代测绘》
2022年第5期38-42,共5页
Modern Surveying and Mapping
基金
江苏省高等学校大学生创新创业训练计划项目(202010298038Z)
江苏省测绘地理信息科研项目(JSCHKY201903)。
关键词
江苏沿海
湿地变化
特征提取
最大似然分类
随机森林分类
Jiangsu coastal area
wetland change
feature extraction
maximum likelihood classification
random forest classification