摘要
准确获取山区植被动态数据对山区生态系统研究及其保护有着重要意义。卫星遥感数据作为获取大尺度山区植被动态的重要数据源,已被制作成各种植被监测产品并被应用于植被绿度、覆盖度以及生产力等研究。然而,不同遥感指数监测山区植被动态及其驱动因子的一致性尚不确定。以尼泊尔地区为例,基于5种遥感指数(MODIS NDVI、MODIS EVI、MODIS LAI、MODIS NPP和OCO-2 GOSIF)和5种气候因子(温度、降水、气压、太阳净辐射和CO_(2)浓度)数据,采用趋势分析和两种残差分析法(多元回归法和一阶偏导法),系统分析了不同遥感指数监测尼泊尔地区植被动态及其驱动因子的不确定性。结果表明:1)2000-2020年尼泊尔地区5种遥感指数均呈现增加趋势,但不同遥感指数空间分布存在差异,MODIS NPP在中山带增加趋势更明显,其他遥感指数在低海拔地区增加态势更明显。2)不同残差分析方法所估算的气候变化对植被变化贡献率差异极大,多元回归法可能严重低估了气候变化的贡献,而一阶偏导法可能高估气候变化的影响。3)基于不同遥感指数的归因结果差异大,气候变化对OCO-2 GOSIF的贡献率显著高于其他指数。研究结果强调了不同遥感指数量化山区植被动态及其原因的巨大分歧,未来需发展更精确的山区植被生长动态检测与归因方法。
Quantifying mountain vegetation dynamics is critical for mountain ecosystem research and protection.As an important data source to obtain large-scale mountain vegetation dynamics,satellite remote sensing data has been extensively used to monitor vegetation greenness,coverage,and productivity.However,the consistency of different remote sensing indices in monitoring of vegetation dynamics and their driving factors in mountainous areas remains unclear.Taking Nepal as an example,this study systematically analyzed the uncertainties of the remotely-sensed vegetation dynamics and driving factors using trend analysis and two residual analysis methods(multiple regression and first-order partial derivative)based on five remote sensing indices(MODIS NDVI,MODIS EVI,MODIS LAI,MODIS NPP and OCO-2 GOSIF)and five climate factors(temperature,precipitation,air pressure,net solar radiation and CO 2 concentration).The results show that:All the five remote sensing indices show increasing trends in Nepal from 2000 to 2020,but the spatial distributions differ by remote sensing index.The increase tendency is more obvious in the middle mountains for MODIS NPP and in low altitude areas for the other indices.The estimated contribution rates of climate change to the vegetation dynamics vary substantially by the two residual analysis methods.The multiple regression method might largely underestimates the contribution of climate change,while the first-order partial derivative method may overestimate the contribution of climate change.The attribution results also differ greatly by remote sensing index,with a clearly larger contribution to OCO-2 GOSIF than the other indices.This study emphasizes the huge discrepancies of remote sensing indices and attribution methods in characterizing the mountain vegetation dynamics and their causes,calling for the development of more accurate methods to detect and attribute the mountain vegetation growth dynamics.
作者
刘子源
周德成
郝璐
樊江文
张良侠
LIU Zi-yuan;ZHOU De-cheng;HAO Lu;FAN Jiang-wen;ZHANG Liang-xia(Key Laboratory of Ecosystem Carbon Source and Sink,China Meteorological Administration(ECSS-CMA),Nanjing University of Information Science and Technology,Nanjing 210044,China;Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China)
出处
《环境生态学》
2023年第8期15-24,共10页
Environmental Ecology
基金
国家自然科学基金项目(42061144004)
国家重点研发计划项目(2021YFB2600100)
第二次青藏高原综合科学考察研究项目(2019QZKK0608)
中国科学院战略性先导科技专项(A类)(XDA20090200)资助。
关键词
遥感
山区
植被动态
不确定性
尼泊尔
Remote sensing
mountain
vegetation dynamics
uncertainty
Nepal
