海岸带信息集成与战略规划研究中心知识库

中国科学院深圳先进技术研究院;环境保护部卫星环境应用中心;中国科学院烟台海岸带研究所;中国科学院南海海洋研究所;

复种指数反映了耕地的实际利用强度,提高农田复种指数是区域粮食增产的重要途径之一,因此,监测和分析复种指数在时间和空间上的变化对粮食安全评估、农业发展规划科学决策有重要意义。NDVI的时间序列蕴涵着植被生长的年循环节律,耕地NDVI时间序列曲线的峰值个数和耕地的种植收割次数相对应,因此耕地的复种指数可通过分析NDVI时间序列曲线来获取。本研究以环渤海地区2000—2009年MODISNDVI时间序列数据为数据源,采用邻域比较法提取耕地NDVI年时间序列曲线的峰值频数,进而计算环渤海地区2000—2009年的复种指数,并对复种指数的时空变化及变化原因进行初步分析。结果显示,在环渤海地区,一年两熟的...

Multiple cropping index (MCI), which is an index for characterizing cropping systems, reflects the degree of arable land available for use at a certain period. It is a significant index for evaluating food production and security and making decisions on agricultural development plans. This is especially useful for China, a country with a large population and smell per-capita arable land. There are two methods (statistical method and remote sensing-based method) for extracting MCI. The second method usually uses Normalized Difference Vegetation Index (NDVI) as source data. NDVI time series for the year can describe the dynamic process of vegetation. For crop, these processes include seeding, jointing, tasseling, harvesting, and so on. Generally, the peak of NDVI time series curves corresponds with tasseling and the lowest point corresponds with harvesting or seeding. Croplands with one crop per year have only one peak and croplands with two crops per year have two peaks. As MCI value matches with the number of peaks of NDVI time series, MCI is extractible from NDVI time series data. In relation to traditional statistical methods, a method based on NDVI time series does not only reflect spatial distribution of MCI, but also easily converge to rapidly provide results. Due to cloud contamination, however, NDVI time series from remote sensing data contain a lot of noise. NDVI data must be preprocessed to remove or reduce noise before extracting for MCI. In this study, the SPLINE interpolation method was used to produce cloud free time series of NDVI to avoid pseudo peaks and accurately extracts MCI from NDVI time series, although not without some limitations. Bohai Rim, an important production base in China, was used as the case-study area in this study. NDVI time series derived from MODIS data were used to extract MCI. MCI for 2000—2009 was extracted and temporal and spatial changes over the Bohai Rim were analyzed. The results showed that croplands in the Bohai Rim with two crops per year mainly occurred in the south of the Great Wall. Other regions of the Bohai Rim were dominated by croplands with one crop per year. The highest MCI was in Shandong Province, where the 10-year mean of MCI was 140.40%. MCI was lowest in Liaoning Province, where natural conditions such as heat were not conducive for producing two crops per year. The 10-year means of MCI for Hebei Province, Tianjin City and Beijing City were 129.65%, 109.52% and 106.13%, respectively. The 10-year mean of MCI for the entire Bohai Rim study area was 117.14%. MCI was also different for different topographic conditions. The mean MCI values for plain, mesa, hill and mountain regions were 154.78%, 117.18%, 109.99% and 103.52%, respectively. In the Bohai Rim, there were obvious inter-annual variations in MCI but with no obvious trends. The maximum MCI was in 2000, while the minimum was in 2009. Inter-annual variation in MCI was mainly influenced by crop rotation and net income from croplands. The existence of mixed pixels affected the accuracy of the extracted MCI based on remote sensing data.