Abstract
Weak spectral response and low model accuracy problems occurred in the process of quantitative inversion of low organic matter content of desert soil in arid areas. This study collects soil samples and field spectral data from different human interference regions in Fukang City, Xinjiang, to search the soil hyperspectral response law that based on the fractional Sprott chaotic system, and combined with the gray system theory to estimate organic matter content rapidly and accurately. Simulation shows that for those sampling soil with lower organic matter content, the range of X and Y components of the dynamic error motion curve distribution of the Sprott chaotic system is larger, and the motion curve of the non-integer order 1.9-order dynamic error is the most obvious. Since the chaotic attractors will appear as a linear trend according to different contents of the organic matter, so this thesis establishes a gray prediction model based on the 1.9 fractional order chaotic attractors. The R 2 , RPD, and RMSE of the low organic matter content in the region without human interruption are 0.995, 14.86, and 0.17, respectively. The R 2 , RPD, and RMSE of low organic matter content in the region with human interruption are 0.992, 11.95, and 0.11, respectively. This study demonstrates that it is feasible to estimate the low organic matter content of desert soils in arid regions via the gray prediction model that based on fractional chaotic attractors. This study provides a novel method for soil spectra signal analysis and estimation of the organic matter content.
Highlights
Soil organic matter (SOM) refers to all carbon-containing organic matter in the soil [1], [2], which can provide nutrients that plants need
The results show that the coefficient of determination of leaf water content (LWC)
This thesis proposes the adoption of fractional-order Sprott chaotic system combined with gray prediction model for prediction
Summary
Soil organic matter (SOM) refers to all carbon-containing organic matter in the soil [1], [2], which can provide nutrients that plants need. Studied the desertification soil in the southeastern part of the Junggar Basin in Xinjiang, and compared the optimum prediction model for quantitative estimation of low organic matter content by adopting multivariate stepwise regression and partial least squares regression (PLSR). Model that based on standard normal variable reflectance (SNV) is the optimum model for the quantitative inversion of the organic matter content of desert soil, and the coefficient of determination of the validation set for sandy soil and clay loam are 0.866 and 0.863, respectively. Hong et al [24] collected the soil from Hubei Province, China, and put forward fractional order derivative to preprocess the hyperspectral reflectance data, and introduced the local modeling technique based on memory learning to estimate the accuracy of SOM, prediction accuracy was compared with partial least square and random forest.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
