Identifying Heavy Metal (Cd) Stress In Rice Using Time-series Signals And Longest Common Subsequence

Heavy metal pollution in farmland is one of the major ecological and environmental problems in the world today,which has the characteristics of wide range,high toxicity and long-term.It will not only reduce the yield of grain,but also seriously threaten human health.Therefore,it is of great significance to use remote sensing technology to quickly,accurately and widely monitor heavy metal stress in crops.Moreover,the growth of rice depends usually on various types of environmental stresses with complex conditions in a large-scale eco-environment system,such as heavy metal stresses,disease,drought,or flood.Specifically,there are different stress types in various spaces,where multiple stresses covered one or several growth stages in the same space.It is also difficult to distinguish heavy metal stress from others in the farmland,due to the similarity of canopy spectral variation induced by multiple stresses.How to distinguish heavy metal stress from other stresses needs further study.In view of the characteristics of "continuous stability" of heavy metal stress,this paper uses the principle and method of time series stability detection to achieve accurate remote sensing recognition and monitoring of heavy metal stress in rice.Taking the rice planting area in Zhuzhou area of Hunan Province as an example,the recognition of cadmium stress(Cd)in rice was studied by acquiring sentinel-2 satellite image data from 2017 to 2019 and combining with the field measured soil heavy metal content data.Firstly,the asymmetric Gaussian function fitting method is used to extract the interval according to the corresponding values of time series points,and then the Gaussian function is used to fit the time series data of each interval.Finally,the leaf area index(LAI)time series curve is obtained through the overall fitting;Then,the improved ensemble empirical mode decomposition(CEEMD)method was used to decompose the Lai time series at multiple scales,and different time series signal components IMF(intrinsic mode function)were obtained;Finally,by solving the longest common subsequence(LCS)between the decomposed time series of the stressed rice and the decomposed time series of the healthy rice,the similarity was calculated,and the stress index was obtained.At the same time,the LCS method was used to measure the similarity between the time series of the adjacent years,that is the interannual change index.The results show that:(1)The LAI time series was decomposed into seven components and one residual term by improved CEEMD,and the combination of the sixth,seventh components and residual term was expressed as the signal of heavy metal stress.(2)The correlation coefficient between stress index and soil heavy metal content was 0.851.The higher the degree of stress,the higher the stress index,and vice versa;In the experimental area,the distribution area proportion of heavy metal stress in rice was relatively low,and mainly concentrated in the west,northeast and Southeast.(3)The inter-annual variation index of the whole experimental area is lower than 0.5,which can be used as an index to measure the stability of heavy metals.The combination of the improved ensemble empirical mode decomposition and the longest common subsequence similarity measure can effectively identify and quantitatively analyze the heavy metal stress status of rice,which provides an important reference for monitoring heavy metal pollution stress of crops.