基于MODIS黄河三角洲湿地NPP与NDVI相关性的时空变化特征

Spatial-temporal variation of NPP and NDVI correlation in wetland of Yellow River Delta based on MODIS data

应用EOS/MODIS卫星遥感资料,分析2001—2006年黄河三角洲地区植被净初级生产NPP及NDVI时空变化特征,在到海岸线不同距离的区域内,讨论了年植被净初级生产力(aNPP)与年均归一化植被指数(ANDVI)的相关性,并建立一元回归分析模型。可以看出,黄河三角洲湿地生态系统分布自海向陆具有明显的空间异质性,滩涂湿地分布于沿海地区,生产力高,该地区aNPP值高且与ANDVI呈显著正相关;其他区域由于是盐碱荒地或农业生态系统,因土壤盐分含量高而系统生产力低,或农耕地受人类干扰严重,一年总的生产力下降,而导致这些区域ANDVI与aNPP呈负相关。利用NDVI进行黄河三角洲生产力监测的可行区域,在距离海岸线小于6_km的范围内,所研究的6a回归模型均通过α=0.01的显著水平的F检验,模型的R2值为0.86—0.97。同时在生产力无法估测的区域,可以使用NDVI有效管理黄河三角洲生态系统,提高其生产力水平。

Study of vegetation productivity and carbon sequestration of wetland ecosystem is a main research topic of global carbon cycle and global change.The ecosystem productivity study of Yellow River Delta,a wetland delta with fastest reclamation speed in China and in the world,is therefore extremely important.Due to human activities have affected all of ecological systems,to find an expeditious way to assess the ecosystem health becomes the major topic of current study.Net primary productivity（NPP） is a key component of terrestrial carbon cycle,and it is defined as the accumulative organic matters by green plants per unit of time and space.NPP,the direct reflection of plant community productivity for a certain natural environment,is the basis of matter and energy cycles of terrestrial ecosystem.The Normalized Difference Vegetation Index has been used for many years to measure and monitor plant growth,vegetation cover,and biomass production from multispectral satellite data.It is an indicator of vegetation growth and its spatial distribution has a close relationship with climate conditions.We use MODIS NPP and NDVI data to analyze the characteristics of spatial-temporal variation of NPP and NDVI in Yellow River Delta between 2001 and 2006.Distribution of NDVI shows that soil water-salt distribution is the main controlling factor of plant growth in the delta.The distribution of primary productivity has been obviously affected by human activities.We further built a new coastal regression model.As the changeable different regions which have different distances from the coastline and which have different lengths,we discuss the correlation between annual NPP（aNPP） and average annual NDVI（ANDVI）.Results suggest that distribution of wetland ecosystems in the Yellow River Delta has obvious spatial heterogeneity from sea to land,along with the ecosystem succession.Mud flat along the coastline is a place for developing ecosystem with a large amount of pioneer plants,which has high productivity and their aNPP is positively related to ANDVI.Saline-alkali land has a low lever of aNPP because of higher concentration of salts;Agricultural land is a developed mature ecosystem in which the nutrition circulates only inside the ecosystem and affected mostly by human activities.Thus there is no directly relationship between aNPP and ANDVI in saline-alkali land and agriculture land,Consequently,we preliminarily discussed the feasible regions in the Yellow River Delta that can use MODIS NDVI to monitor their productivity.In the region where the distance to coastline is less than 6km,all of six simple regression models for each year passed the significant F-test（α=0.01,R2 0.86 — 0.97）.Finally,to understand the time differentiation characteristics of correlation coefficient between ANDVI and aNPP,we calculated the percentage size and the average aNPP with different ANDVI value range.We figured out that resource competition may be one of the fact making the correlation coefficient difference.Therefore,we can use NDVI to effectively monitor and manage the ecosystems and to increase their productivity.Based on the research of the ecosystem productivity of Yellow River Delta,we can effectively understand the regularity for change of productivity during the rapid economic development period,which can further guide the administration of local ecosystems.