基于遥感数据的江苏省水稻收获期时空分析研究

本研究基于2001—2020年中分辨率成像光谱仪遥感产品,使用谐波分析重建完整作物生长过程植被指数变化曲线,对江苏省范围内水稻种植区域的收获期进行提取。同时利用Sen趋势分析与相关分析方法,研究水稻收获期时空动态变化趋势与影响因子。结果表明:利用谐波分析方法能够较好地重建作物生长期生理参数变化曲线,滤波值与原始值线性相关系数达到0.8;水稻收获期分布在9月中旬到12月上旬之间,2010年之后,10月中下旬之后占比增加,达到50%以上;部分地区水稻收获期呈现延迟趋势,南京北部、扬州、盐城北部地区为3 d·a^-1及以上,泰州、南通等地区保持稳定,趋势绝对值低于1 d·a^-1;9—11月降水日数与降水量均对水稻收获期有影响,10月降水量与收获期存在负相关,10、11月降水日数与收获期存在正相关。本研究的基于遥感的水稻收获期提取方法为大范围的水稻农业物候信息提取提供参考。

Spatial and temporal analysis of rice harvesting period in Jiangsu Province based on remote sensing data

Based on MODIS remote sensing products from 2001 to 2020, a complete crop growth process vegetation index change curve was reconstructed by using harmonic analysis to extract the harvest period of rice-growing areas within Jiangsu Province. By using the Sen trend analysis and correlation analysis methods, the spatial and temporal dynamics of the rice harvesting period were studied in terms of trends and influencing factors. Results show that the harmonic analysis method is able to reconstruct the change curves of physiological parameters of the crop during the growing period, and the filtered values reach a linear correlation of 0.8. The rice harvesting period is distributed between mid-September and early December, and after 2010, moreover, the proportion increases after mid-to-late October, reaching more than 50%. The rice harvest period in some areas shows a delayed trend, with 3 days/year and more in northern Nanjing, Yangzhou, and northern Yancheng, but remaining stable in Taizhou and Nantong, where the absolute value of the trend is less than 1 day/year. Both the precipitation days and the precipitation amount from September to November affects the rice harvest period, with a negative correlation between precipitation and harvest period in October and a positive correlation between precipitation days and harvest period in October and November. The findings of the study can provide support for the extraction and analysis of agrometeorological information on rice on a large scale.