| Arable land is the essence of land resources,which is a basic condition for human access to agricultural products.Multiple cropping index is a very important indicator in agricultural statistic in China and it represents the situation of arable land use effectively.In recent years,remote sensing data has been widely applied in extraction of multiple cropping index(MCI),however,the clouds,aerosols and snow et al.sometimes influence the data quality potentially.Therefore,NDVI time-series data need to be smoothed before being used.Although many methods have been developed to de-noise time-series NDVI data and to extraction of MCI)based on these reconstructed long timeseries datasets,there are few studies focused on the effectiveness and the extraction accuracy of these techniques systematically.In this research,the smooth crops growth NDVI curves from 2001 to 2013 were rebuilt by the different smoothing techniques.Based on the smoothed data and field investigation data,we constructed a method which was suitable for the extraction of MCI in the Poyang Lake Plain,and then set up a dataset of MCI in the plain of Poyang Lake.Based on the dataset,we analyzed spatial and temporal changes of MCI from 2001 to 2013 in oyang Lake region and its driving factors.The results showed that:(1)All of the de-noising techniques were effective for reconstructing high-quality NDVI time-series data sets,while the S-G performed best in the high vegetation coverage region;(2)The multiple cropping index of cultivated land using all methods had higher precision,but there were differences among different methods.the S-G technique performed best among six kind of methods;(3)Double-or triple-cropping rotation systems were commonly applied in Poyang Lake Plain during 2001-2013,while single cropping system was mainly found in Xingzi,Anyi,Xinjian,Poyang and Xinyu;(4)The MCI in Poyang Lake Plain showed an overall ascending trend during 2001-2013 with inter-annual fluctuations and exhibited significant spatial heterogeneity.Aside from Ruichang,Jiujiang,Zhangshu,Fengcheng,Jinxian,Fuzhou,Dongxiang and Yingtan,all counties displayed an increasing trend in MCI from 2001 to 2013,particularly in Yongxiu,Anyi,Poyang and Yugan(p<0.1);(5)Areas where the MCI decreased then increased and areas with a continual increasing trend occupied 29.48% and 28.63% of the total cultivated land respectively,followed by areas with a continual decreasing trend(23.2%).Areas where the MCI increased then decreased accounted for the smallest proportion;(6)The average mechanical power of cultivated land,average number and total wagesof fully employed staff and workers in urban area were the major factors which influenced the multiple cropping index in the Poyang Lake Plain,but there were some differences in the influencing factors among different regions. |