| Multiple cropping is an important way of cultivation in China, it also has a vital significance on maintaining and increasing the grain production and promoting economic development in rural area. The farmland cropping system in our country is complicated and diverse. In the past thirty years, the farmland resource in our country had a complex changes in quantity, quality and use structure due to the influence caused by natural conditions and rural social conditions, which requires us to be able to have a good grasp of the situation of the farmland resource and the food production in China in order to make scientific decisions on estimating the changes and its reasons of grain production.Taking China as the research area, this paper used the GLASS LAI data provided by Beijing Normal University and used the TIMESAT software to reconstruct the time series GLASS LAI data to get time series GLASS LAI smooth curve which can reflect the different cropping system. Based on the R language program to seek the maxima numbers on a smooth curve(that is the numbers of the peak of wave), the automatic extraction of the information of the farmland cropping system in China was realized. After using farmland cover data to cut the map, the distribution map of Chinese farmland cropping system in 1982, 1992, 2002 and 2012 were obtained. According to the four maps, this paper analyzed the changes of the farmland cropping system in China in recent thirty years, calculated the multiple-cropping index and compared the results obtained separately with the statistics and the previous research’s results. The main research methods and conclusions are as follows:(1) After the preprocessing of the totally twelve years’ time series GLASS LAI data of Chinese 32 provinces, the time series GLASS LAI curve was established. But there was noise in the image, the time series GLASS LAI curve presented jaggies occasionally. The TIMESAT 3.0 software operated under the Matlab provided the Gaussian filter, the Logistic filter and the Savitzky-Golay filter(short for the S-G filter) to reconstruct the curve. The result showed that, the Gaussian filter and the Logistic filter were not fitting the curve of two and three ripe crops well, but the S-G filter were better to fit the curve of different cropping system. It could present the seasonal changes of crops, which was help to extract the number of the peak of wave.(2) According to the changing characteristics of the time series curve formed by the phonological characteristics of plants, cropping system was equal to the number of the maximum points on curves. After saving the time series GLASS LAI smooth curves which reflected the different cropping system as text format, used the R language program to extract the maximum numbers which met conditions on a smooth curve. The results showed that this method could extract the numbers of peaks on time series smooth curves for each pixel, with 0 representing noncropland, 1 representing single cropping system, 2 representing of two cropping system, 3 representing triple-cropping system.(3) From the distribution map finally got of Chinese farmland cropping system in 1982, 1992, 2002 and 2012, we could clearly see that our country’s cropping system was from single to complex with the area from north to south, meanwhile from single cropping system to two cropping system and then triple-cropping system. The areas of single and two cropping systems were most concentrative, but the areas of triple-cropping system were scattered with partly centralized. In recent thirty years, the changes of Chinese farmland cropping system were as follows: single cropping system areas’ overall changes were not obvious, but there were changes in parts of regions, some parts increasing and some decreasing. Two cropping system areas increased significantly, with the extending from Henan to all around. Triple-cropping system areas were sporadic and expanding slowly.(4) The attribute information of the Arc GIS mapping provided the numbers of pixels of different cropping system. This paper used this data to calculate the multiple-cropping index of Chinese 32 provinces. The results showed that over the past three decades, Chinese multiplecropping index were rising constantly which indicated that more and more farmland turned to two cropping system or triple-cropping system. At last, this paper compared the results of 2002 separately with the statistics and the previous research’s results. The comparison showed that the results of this method and statistics or previous research’s results were a positive correlation, the correlation coefficients were 0.788 and 0.903, which verified the reliability of this study. |
PDF Full Text Request