点击蓝字 关注我们
今天给大家分享的论文是兰州大学地球与环境科学学院科研团队在Remote Sensing期刊上发表的《Land Cover Changes of the Qilian Mountain National Park inNorthwest China Based on Phenological Features and Sample Migration from 1990 to 2020》。本研究旨在基于西北地区的祁连山国家公园(QMNP)样本迁移方法生成土地覆盖分类的长时间序列数据集。将 Landsat 5、7 和 8 影像和野外样本数据与多个影像要素和随机森林算法相结合,完成了 1990 年至 2020 年 QMNP 的土地覆被分类。
论文信息
题目:Land Cover Changes of the Qilian Mountain National Park inNorthwest China Based on Phenological Features and Sample Migration from 1990 to 2020
关键词:Landsat;sample migration;Jeffries–Matusita distance;machine learning
作者:Yanyun Nian, Zeyu He, Wenhui Zhang and Long Chen
DOI:https://doi.org/10.3390/rs15041074
论文摘要
The spatial and temporal variation analysis of land cover classification is important forstudying the distribution and transformation of regional land cover changes.The Qilian MountainNational Park(QMNP),an important ecological barrier in northwestern China,has lacked land coverproducts for long time series.The Landsat images available on the Google Earth Engine(GEE)makeit possible to analyze the land cover changes over the past three decades.The purpose of this study was to generate a long time series of datasets of land cover classification based on the method ofsample migration in the QMNP in Northwest China.The Landsat 5,7,and 8 images and field sampledata were combined with multiple image features and the random forest algorithm to complete theland cover classification of the QMNP from 1990 to 2020.The results indicate that(1)the method ofJeffries–Matusita(J-M)distance can reduce image feature redundancy and show that elevation andphenological features have good differentiability among land cover types that were easy to mix withfeature classes;(2)the spatial distribution of land cover every 10 years between 1990 and 2020 wasconsistent in the QMNP,and there were obvious differences in land cover from the east to the westpart of the QMNP,with a large area of vegetation distribution in Sunan county in the central partand Tianzhu county in the east part of the QMNP;(3)over the past 30 years,forests and grasslandsdecreased by 62.2 km2and 794.7 km2,respectively,while shrubs increased by 442.9 km2in the QMNP.The conversion of bare land to grassland and the interconversion between different vegetation typeswere the main patterns of land cover changes,and the land cover changes were mainly concentratedin pastoral areas,meaning that human activity was the main factor of land cover changes;and(4)when the samples of 2020 were migrated to 2010,2000,and1990,the overall classificationaccuracies were 89.7%,88.0%,86.0%,and 83.9%,respectively.The results show that the vegetationconservation process in the QMNP was closely related to human activities.
土地覆被分类的时空变化分析对于研究区域土地覆被变化的分布和变换具有重要意义。祁连山国家公园 (QMNP) 是中国西北地区重要的生态屏障,长期以来一直缺乏土地覆盖产品。Google Earth Engine (GEE) 上提供的 Landsat 影像可用于分析过去三十年的土地覆被变化。本研究旨在基于西北地区 QMNP 样本迁移方法生成土地覆盖分类的长时间序列数据集。将 Landsat 5、7 和 8 影像和野外样本数据与多个影像要素和随机森林算法相结合,完成了 1990 年至 2020 年 QMNP 的土地覆被分类。结果表明:(1)Jeffries-Matusita(J-M)距离方法可以减少影像特征的冗余,表明高程和物候特征在容易与要素类混合的土地覆盖类型之间具有良好的可区分性;(2)1990—2020年QMNP每10年土地覆被空间分布一致,且QMNP东西区土地覆被分布差异明显,中部肃南县和东部天祝县植被分布面积较大;(3) 近 30 年来,森林和草原减少了 62.2 平方公里和 794.7 平方公里,而灌木增加了 442.9 平方公里。在 QMNP 中。裸地向草原的转换和不同植被类型之间的相互转换是土地覆被变化的主要模式,土地覆被变化主要集中在牧区,即人类活动是土地覆被变化的主要因素;(4) 将 2020 年的样本迁移到 2010 年、2000 年和 1990 年时,总体分类准确率分别为 89.7%、88.0%、86.0% 和 83.9%。结果表明:QMNP 植被保护过程与人类活动密切相关
重要图表
——研究区域的概况图——
——GEE 上的样本迁移、物候特征提取和土地覆盖分类的工作流程——
——2020年样本空间分布——
——土地覆被分类精度随决策树数量的变化——
——2020 年 Landsat 8 不同土地覆被类型之间的 J-M 距离——
——土地覆被分类精度随要素组合的变化——
——2020 年的土地覆被分类结果——
本文创新
本文使用Jeffries-Matusita(J-M)距离方法有效减少影像特征的冗余,表明高程和物候特征在容易与要素类混合的土地覆盖类型之间具有良好的可区分性。
心得收获
使用 J-M 距离可以很好地优化特征集,将 35 个特征减少到 29 个,整体精度从 88.6% 提高到 89.7%。
土地覆盖变化的结果可以揭示 QMNP 的哪些区域在植被覆盖度(如森林和灌木丛)方面发生了最大的变化,从而转移工作重心。
欢迎指出文中翻译存在不准确的地方!
微信公众号
@Study and Share
抖音
@生态环境遥感监测研究中心
小红书
@生态环境遥感监测研究中心