A Modified Shape Model Incorporating Continuous Accumulated Growing Degree Days for Phenology Detection of Early Rice
A Modified Shape Model Incorporating Continuous Accumulated Growing Degree Days for Phenology Detection of Early Rice
复制标题
DOI：
10.3390/rs14215337
复制
发表时间：
2022-10
期刊：
影响因子：
--
通讯作者：
Shicheng Liao;Xiong Xu;Huan Xie;Peng Chen;Chao Wang;Yanmin Jin;X. Tong;C. Xiao
中科院分区：
文献类型：
--
作者：
Shicheng Liao;Xiong Xu;Huan Xie;Peng Chen;Chao Wang;Yanmin Jin;X. Tong;C. Xiao
利用形状模型（SM）是利用长时间序列卫星遥感数据确定作物物候期的典型方法。基于AGDD的平均形状模型（AAGDD-SM）与SM相比考虑了温度，但是，由于全天温度可能发生重大变化，常用的日平均温度不足以确定准确的AGDD。本文提出了一种改进的形状模型，用于更好地估计物候期，并将其纳入连续AGDD（CAGDD）中，后者是基于一天内连续24 h的温度计算的，不同于日历日或平均AGDD指标。在这项研究中，CAGDD取代了中国江西省早稻试验点5年（2014年至2018年，不包括2015年）的NDVI生长曲线的横坐标。根据田间物候观测资料，确定了返青期、分蘖期、抽穗期和开花期4个关键物候期。结果表明，与AAGDD-SM相比，本文提出的方法对各物候期的预测效果基本得到改善。对于平均气温低于最低气温（K1）但有效积温不为零的情况，按本文的方法可以计算出较精确的AGDD。
Using a shape model (SM) is a typical method to determine the phenological phases of crops with long-time-series satellite remote sensing data. The average AGDD-based shape model (AAGDD-SM) takes temperature into account compared to SM, however, the commonly used daily average temperature is not sufficient to determine the exact AGDD owing to the possibly significant changes in temperatures throughout the day. In this paper, a modified shape model was proposed for the better estimation of phenological dates and it is incorporated into the continuous AGDD (CAGDD) which was calculated based on temperatures from a continuous 24 h within a day, different from the calendar day or the average AGDD indicators. In this study, the CAGDD replaced the abscissa of the NDVI growth curve over a 5-year period (2014 to 2018, excluding 2015) for a test site of early rice in Jiangxi province of China. Four key phenological phases, including the reviving, tillering, heading and anthesis phases, were selected and determined with reference to the field-observed phenological data. The results show that compared with the AAGDD-SM, the method proposed in this paper has basically improved the prediction of each phenological period. For those cases where the average temperature is lower than the minimum temperatures (K1) but the effective accumulated temperature is not zero, more accurate AGDD can be calculated according to the method in this paper.