Comparison of Red Edge Parameters of Winter Wheat Canopy under Late Frost Stress
WU Yong-feng1, HU Xin2, Lü Guo-hua1, REN De-chao2, JIANG Wei-guo3, SONG Ji-qing1*
1. Key Laboratory of Agricultural Environment，Ministry of Agriculture，Institute of Environment and Sustainable Development in Agriculture，Chinese Academy of Agricultural Sciences，Beijing 100081，China 2. Wheat Research Laboratory, Shangqiu Academy of Agriculture and Forestry Sciences，Shangqiu 476000，China 3. State Key Laboratory of Earth Surface Processes and Resource Ecology，Beijing Normal University，Beijing 100875，China
Abstract：In the present study, late frost experiments were implemented under a range of subfreezing temperatures (-1～-9 ℃) by using a field movable climate chamber (FMCC) and a cold climate chamber, respectively. Based on the spectra of winter wheat canopy measured at noon on the first day after the frost experiments, red edge parameters REP, Dr, SDr, Drmin, Dr/Drmin and Dr/SDr were extracted using maximum first derivative spectrum method (FD), linear four-point interpolation method (FPI), polynomial fitting method (POLY), inverted Gaussian fitting method (IG) and linear extrapolation technique (LE), respectively. The capacity of the red edge parameters to detect late frost stress was explicated from the aspects of the early, sensitivity and stability through correlation analysis, linear regression modeling and fluctuation analysis. The result indicates that except for REP calculated from FPI and IG method in Experiment 1, REP from the other methods was correlated with frost temperatures (P<0.05). Thereinto, significant levels (P) of POLY and LE methods all reached 0.01. Except for POLY method in Experiment 2, Dr/SDr from the other methods were all significantly correlated with frost temperatures (P<0.01). REP showed a trend to shift to short-wave band with decreasing temperatures. The lower the temperature, the more obvious the trend is. Of all the REP, REP calculated by LE method had the highest correlation with frost temperatures which indicated that LE method is the best for REP extraction. In Experiment 1 and 2, only Drmin and Dr/Drmin calculated by FD method simultaneously achieved the requirements for the early (their correlations with frost temperatures showed a significant level P<0.01), sensitivity (absolute value of the slope of fluctuation coefficient is greater than 2.0) and stability (their correlations with frost temperatures always keep a consistent direction). Dr/SDr calculated from FD and IG methods always had a low sensitivity in Experiment 2. In Experiment 1, the sensitivity of Dr/SDr from FD was moderate and IG was high. REP calculated from LE method had a lowest sensitivity in the two experiments. Totally, Drmin and Dr/Drmin calculated by FD method have the strongest detection capacity for frost temperature, which will be helpful to conducting the research on early diagnosis of late frost injury to winter wheat.
[1] YANG Bang- jie，WANG Mao-xin，PEI Zhi-yuan(杨邦杰，王茂新，裴志远). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报)，2002，18(2)： 136. [2] Smith K L，Steven M D，Colls J J. Remote Sensing of Environment，2004，92(2)： 207. [3] Clevers J G P W，Kooistra L，Salas E A L. International Journal of Remote Sensing，2004，25(19)： 1. [4] WANG Hong-bo，FENG Rui，JI Rui-peng，et al(王宏博，冯 锐，纪瑞鹏，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析)，2012，32(12)： 3358. [5] LI Zhang-cheng，ZHOU Qing-bo，Lü Xin，et al(李章成，周清波，吕 新，等). Acta Agronomica Sinica(作物学报)，2008，34(5)： 831. [6] Guyot G, Baret F, Jacquemoud S. Imaging Spectroscopy for Vegetalion Studies. in Imaging Spectroscopy: Fundamentals and Prospective Application, Edited by Toselli F, Bodeohtel J. Dordrecht: Kluwer Academic Publishers, 1992. 145. [7] Miller J R, Hare E W, Wu J. International Journal of Remote Sensing, 1990, 11(10): 1755. [8] Dawson T P, Curran P J. International Journal of Remote Sensing, 1998, 19(11): 2133. [9] Pu R, Gong P, Biging G S, et al. IEEE Transaction on Geoscience and Remote Sensing, 2003, 41(4): 916. [10] Cho M A, Skidmore A K. Remote Sensing of Environment, 2006, 101(2): 181. [11] Gong P，Pu R，Heald R C. International Journal of Remote Sensing，2002，23(9)： 1827. [12] WANG Yuan-yuan，CHEN Yun-hao，LI Jing，et al(王圆圆，陈云浩，李 京，等). Journal of Remote Sensing(遥感学报)，2007，11(6)： 875. [13] ZHU Yong-li，LI Ping-ping，MAO Han-ping，et al(朱咏莉，李萍萍，毛罕平，等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报)，2011，42(11)： 174. [14] Carter G A，Miller R L. Remote Sensing of Environment，1994，50(3)： 295. [15] Zhong X L，Mei X R，Li Y Z，et al. Journal of Agronomy and Crop Science，2008，194(5)： 343.
