文献类型：期刊文献

中文题名：环塔里木盆地1990-2021年改进型土壤调整植被指数时空数据集

英文题名：A dataset of spatial and temporal modified soil adjusted vegetation indices in the Ring Tarim Basin from 1990 to 2021

作者：戚曌[1,2,3] 席磊[1,2,3] 曹晓明[1,2,3] 崔梦淳[1,2,3] 邹佳秀[1,2,3] 刘洪霞[4,5]

第一作者：戚曌

机构：[1]中国林业科学研究院生态保护与修复研究所,北京100091;[2]中国林业科学研究院荒漠化研究所,北京100091;[3]荒漠生态系统与全球变化国家林业和草原局重点实验室,北京100091;[4]中国农业科学院农业信息研究所,北京100081;[5]农业部农业大数据重点实验室,北京100081

年份：2025

卷号：10

期号：2

起止页码：506-514

中文期刊名：中国科学数据(中英文网络版)

外文期刊名：China Scientific Data

基金：第三次新疆综合科学考察课题(2021xjkk0304)。

语种：中文

中文关键词：环塔里木盆地;改进型土壤调整植被指数;线性回归;1990-2021

外文关键词：the Ring Tarim Basin;MSAVI;linear regression;1990-2021

分类号：P23

摘要：塔里木盆地是我国荒漠化问题较为严重的区域之一,生态环境脆弱,受人为活动影响较大。植被是影响地区荒漠化程度的关键,可以作为监测区域荒漠化状况的考虑因素,对于植被稀疏的干旱地区常采用改进型土壤调整植被指数(Modified Soil Adjusted Vegetation Index,MSAVI)来反映其分布状况。研究以1990-2021年的Landsat影像为数据源,基于谷歌地球引擎(Google Earth Engine,GEE)云平台,采用最大值合成法合成环塔里木盆地生长季(4-10月)数据影像,并计算得到环塔里木盆地1990-2021年间7个时期的MSAVI数据集。研究获得了研究区122块实测的样地FVC数据用来做验证样本,然后基于MSAVI采用像元二分法计算得到植被覆盖度,分别提取以样本点为核心的1×1和3×3像元范围的均值作为计算得到的植被覆盖度值,来消除实测样本点坐标范围可能带来的偏差,得到的样本精度分别为:R2=0.737,R2=0.760,均方根误差分别是RMSE=0.125,RMSE=0.124,间接验证了本数据集的质量以及MSAVI指数反演植被覆盖度的可行性。本数据集可为干旱区植被状况以及荒漠化程度变化等方面的了解提供数据支撑。
The Tarim Basin is one of the serious desertification areas in China with a fragile ecological environment that is highly affected by human activities.Vegetation is the key factor in determining the degree of desertification in a region and can serve as an important factor for monitoring regional desertification status. The Modified Soil Adjusted Vegetation Index (MSAVI) is often used to reflect the distribution of vegetation in arid areas with sparse vegetation. Based on the data sources of Landsat images from 1990 to 2021Landsat images from 1990 to 2021, this study used the Google Earthe Engine (GEE) cloud platform, and applied the maximum value synthesis method to synthesize data images for the growing season (April to October) in the Ring Tarim Basin, resulting in a dataset of MSAVI during seven periods between 1990 and 2021. The study obtained FVC data from 122 measured sample plots in the study area for use as verification samples. Then, using the MSAVI indexes, vegetation coverage was calculated through a pixel dichotomy inversion method. The mean values of vegetation coverage within 1×1 and 3×3 pixel ranges centered on each sample point were extracted respectively, so as to eliminate potential deviations caused by the coordinate ranges of the measured sample points. The results showed R2 values of 0.737 and 0.760, with RMSE values of 0.125 and 0.124, respectively. These results verified the quality of this dataset and the feasibility of using the MSAVI index for vegetation coverage inversion. This dataset can provide data support for understanding vegetation status and desertification levels in arid areas.