Dynamic monitoring of multiple cropping index of paddy field based on MODIS-EVI data in Guangdong province

Due to the very fragmentation of cultivated land in Guangdong Province, there are few studies on the cultivated land multiple cropping index (MCI), which can not meet the needs of agricultural production and policy. Therefore, this paper selects the 2015 paddy field data of Guangdong Province and uses the 16-day synthetic MODIS-EVI data from 2014 to 2016, Savitzky-Golay and Asymmetric Gaussian methods are used to reconstruct multi-temporal remote sensing data, and quadratic difference algorithm is used to extract the paddy field MCF. A comparative analysis of the two methods shows that the Asymmetric Gaussian function fitting is more suitable for a single season. The Savitzky-Golay filtering is more sensitive, and there are many pseudo-peaks in the fitted curve, resulting in a large extraction result compared with verification data. The twice Savitzky-Golay filtering further smooths the curve and removes a large number of false peaks, which is more suitable for the vegetation characteristics of paddy fields in Guangdong Province; The paddy field planting area is highly correlated with the rice planting area, but there are rice-peanut, rice-sweet potato and rice-sweet sugarcane planting patterns in the paddy field. In addition, the classification accuracy of paddy fields is one of the main influencing factors, so it is difficult to extract rice planting area accurately; During 2014-2016, the paddy field in Guangdong Province is dominated by double cropping system. The area of single cropping system is increased and then decreased, and the area of the double cropping system is reduced and then increased. The fallow paddy fields are mainly distributed around the construction land, especially in the Guangdong-Hong Kong-Macao Greater Bay Area.