Inversion And Mapping Of Soil Organic Matter In Shrimp And Rice Area Based On The Feedback Spectrum Of Soil Spectral During The Crop Growth Period

The description of the spatial distribution and time-varying characteristics of soil attributes can effectively guide humans to use and protect soil more effectively and reasonably.Digital soil mapping is an important means to describe soil properties intuitively.In the vast plain area,the lack of sufficient terrain factors to guide soil mapping will significantly reduce the accuracy of digital soil mapping.Professor Zhu Axing first proposed the concept of soil spectral feedback surface,using the change characteristics of soil surface spectrum in consecutive days after precipitation events to describe soil properties.Although this method can improve the accuracy of digital soil mapping,the conditions for obtaining remote sensing data are harsh.It needs to screen a large number of remote sensing images.This paper selects Yunjiang Lake Farm in Qianjiang City on the Jianghan Plain as the research area to explore the impact of soil spectral feedback on the whole organic growth cycle of rice in the digital mapping of soil organic matter.This study set up three groups of 62 sampling points.The sampling time is June,July,August and October,including the full growth period of rice.Record the coordinates of each point,soil samples of rice fields and shrimp trenches,aerial photography data,vegetation photos,soil hyperspectral data and remote sensing image data.The main contents of the research are the following three points:First,the differences in organic matter content,soil particle size,soil pH and soil hyperspectral data between the paddy field soil and shrimp ditch soil under the shrimp-rice co-cultivation model;During the period,the organic matter,particle size,pH,hyperspectral data,and normalized vegetation index of paddy field and shrimp ditch were changed.Third,the remote sensing images of 13 bands of Sentinel 2 satellite were used to establish the soil spectrum of the whole growth cycle of rice Feedback surface.Analyze the effect of the soil spectral feedback surface of the whole growth cycle on the digital mapping of soil organic matter.Overall,the soil organic matter content in the study area generally showed a decreasing trend from northwest to southeast;the soil pH range was generally in a neutral state;the difference in soil particle size was small,but in the northwest and south-central of the study area There are areas where the soil particle size is relatively large.The model established by the environmental covariate system including the characteristic variables of the soil spectral feedback surface has a root-mean-square error of approximately 5.87 and a determination coefficient of approximately 0.72 for the estimation of soil organic matter.The model of environmental covariate system without soil spectral feedback surface has an rms error of 6.14 and a determination coefficient of 0.70 for the estimation of soil organic matter.After analyzing the experimental content and results,the soil spectral feedback surface during the whole growth period can be used as a new environmental covariate for digital soil mapping in plain areas.The feature vector extracted from the soil spectral feedback surface can significantly improve the digital mapping of soil organic matter.Accuracy;the remote sensing data using the Sentinel 2 satellite can effectively support digital soil mapping,and the environmental covariates calculated based on its remote sensing data show a high importance;under the shrimp-rice co-crop model,the paddy soil and shrimp ditch The soil has obvious differences in soil organic matter,but the differences in pH and particle size are not obvious.