Monitoring variations in vegetation net primary productivity (NPP) contributes greatly to the evaluation of terrestrial ecosystem structure and function. Current research has focused on the trend changes of vegetation productivity but has largely ignored spatial correlation. This study selected MOD17A3 NPP data in Inner Mongolia to explore the spatiotemporal variations of NPP using spatial autocorrelation indices (Global Moran’s I, Getis-Ord General G, Getis-Ord G_i*, and Anselin Local Moran’s I). This is helpful to examine whether NPP in Inner Mongolia has exhibited significant spatial cluster based on temporal variations and spatial patterns. NPP in Inner Mongolia has shown significant spatial cluster in high-value areas from 2000 to 2014 and demonstrated a polarized distribution. Although the east part of the study area generally showed a high NPP pattern, the northeast turned to low NPP pattern in 2001, 2004, 2007, and 2010 and the southeast turned to low NPP pattern in 2009 and 2013. The former variations were primarily affected by precipitation, while the latter variations were affected by both hydrothermal conditions and human activities. These results are useful to analyze spatial correlations and local abnormalities of NPP and provide references for integrated ecosystem management in Inner Mongolia.

监测植被净初级生产力（NPP）的变化极大地有助于评估陆地生态系统的结构和功能。当前的研究集中在植被生产力的趋势变化上，但很大程度上忽略了空间相关性。这项研究选择了内蒙古的MOD17A3 NPP数据，以利用空间自相关指数（Global Moran's I，Getis-Ord General G，Getis-Ord G _i*和Anselin Local Moran的I）。这有助于研究内蒙古的核电厂是否根据时间变化和空间格局表现出显着的空间集群。内蒙古的核电厂从2000年到2014年在高价值地区显示出显着的空间集群，并呈现出极化分布。尽管研究区域的东部总体上呈现出较高的NPP格局，但东北部在2001年，2004年，2007年和2010年转向了低NPP格局，东南部在2009年和2013年转向了低NPP格局。前者的变化主要受到影响受降水影响，而后者的变化受热液条件和人类活动的影响。这些结果可用于分析核电厂的空间相关性和局部异常，为内蒙古生态系统综合管理提供参考。