Extraction of paddy rice planting areas in county using both GEE and multi-source Sentinel data
As one of the main food crops for human beings,paddy rice plays a vital role in meeting the growing food demand in the world. Accurate and efficient access to the paddy rice planting area is a key basis for decision-making and resource allocation. However,how to efficiently extract the paddy rice planting area using multi-source remote sensing data in a complex farmland environment is still a challenging task. This paper took paddy rice in Gaoan City as the research object. Based on the Google Earth Engine (GEE) platform,combined with Sentinel-2 optical imagery and Sentinel-1 synthetic aperture radar (SAR) data,this paper constructed a variety of classification features according to different phenological stages of paddy rice and developed an effective extraction method for paddy rice planting area. The normalized difference vegetation index (NDVI) and vertical-horizontal polarization (VH) time series curves of different vegetation were analyzed emphatically by the harmonic fitting model. The results show that the fitted NDVI and VH time series of paddy rice have unique morphological characteristics. The fusion of multi-source Sentinel data can effectively improve the classification accuracy,and the overall accuracy and Kappa coefficient are 96.49％ and 95％,respectively. The results can provide strong support for the sustainable development of local agriculture and resource management and provide a scientific basis for stabilizing the production area of double-cropping rice and ensuring food security.
paddy ricetime series curvesentinel datarandom forest (RF)
万方数据
维普
