Study On Soil Moisture Monitoring Method Of Winter Wheat In Large Area Based On Modified Vegetation Index

Soil moisture refers to the state of surface soil moisture,which is an important indicator parameter in the fields of hydrology,meteorology,and agriculture science.Henan is an important producing area in China,and its production is susceptible to drought.Monitoring soil moisture in a timely and accurate manner is of great significance for ensuring grain production.The traditional soil moisture measurement methods such as field measurement are accurate,however,they cannot achieve dynamic real-time monitoring of large-scale farmland.Remote sensing technology can make up for the shortcomings of traditional methods.Research on remote sensing monitoring of soil moisture has tremendous practical significance.Vegetation indexbased monitoring method is a common method for remote sensing monitoring,but the problem of low monitoring accuracy generally exists.In addition,there are some problems like more spot-shaped areas but less macro-major areas studies,more time cross-sectional but less long-term growth periods studies.Those problems limit the further application of this method.For the above problems,this paper takes the winter wheat farmland moisture content in the Huanghuaihai Plain in Henan Province as the research object.MODIS and Landsat8 remote sensing images and measured moisture content are used to modify or integrate the vegetation index according to the actual conditions of different growth periods to improve the accuracy of soil moisture monitoring and inversion.The main contents include:(1)Soil moisture monitoring based on modified NDVI.NDVI is the earliest and most widely used vegetation index,firstly select NDVI for experiments.The growth period of winter wheat in the Central Plains is as long as eight months,and ecological factors such as light,heat,water,and soil are complex and changeable.Considering the differences in light and heat conditions and crop phenological periods in large regions,this paper divided the planting area and growth period(based on wheat phenology polymerization).In each area and growing period,this paper corrects NDVI and monitor soil relative humidity according to the temperature,spectrum and other factors.The experimental results show that,in general,the relationship between modified NDVI and measured soil moisture is better than that of NDVI,and the accuracy of monitoring and inversion of soil moisture is improved to some extent.However,the sensitivity to modified NDVI is different in different growth periods,poor in early period.Experiments have proved that the soil moisture monitoring method based on the modified vegetation index is feasible.(2)Analysis of the response relationship between multiple vegetation indices and soil moisture.Based on the above experiments,considering the response relationship of different indices to soil moisture at different depths.According to the actual situation of the data source,this paper select Normalized Difference Vegetation Index(NDVI),Enhanced Vegetation Index(EVI),Vegetation Supply Water Index(VSWI)and Temperature Vegetation Drought Index(TVDI),analyze the correlation between them and soil moisture at different depths after the regreening period,establish regression models and analyze the prediction accuracy.The results show that TVDI is the optimal response index,and 10-20 cm soil is the optimal response depth.(3)Research on adaptive soil moisture monitoring method based on modified or integrated vegetation index method.Taking the vegetation coverage into consideration,regarding the start date of the jointing period as the demarcation,the growth period of winter wheat is divided into the early-growth period(low vegetation coverage period from sowing to regreening)and the later-growth period(full vegetation coverage period from jointing to harvesting period).According to the characteristics of the period,different adaptive methods were used to monitor the soil moisture.The early-growth period of winter wheat is in the vegetative growth stage.From bare soil to semivegetated coverage,vegetation coverage is increasing.Therefore,considering the two factors of soil and vegetation,the ATI and NDVI feature spaces are established.Based on this,the VADI is proposed.Using VADI to monitor the soil moisture content of 10-20 cm,compared the result with ATI and NDVI.The results show that the correlation coefficients between VADI and soil moisture are on average 0.15 and 0.07 higher than NDVI and ATI.The standard deviations of moisture inversion are 0.87 and 0.53 lower respectively.The VADI is better.The later-growth period of winter wheat is in the reproductive growth stage,the vegetation coverage is high and remained largely unchanged.When the optimal vegetation index(TVDI)to inverse soil moisture is used,the non-vegetated areas of mixed pixels affect the accuracy of vegetation index calculation.For this problem,Landsat8 remote sensing data is used to calculate the vegetation coverage factor with higher spatial resolution,and the physical statisticsbased method is used to fuse with MODIS data,and then TVDI is modified.Using modified TVDI to monitor the soil moisture content and compared the result with TVDI.The results show that,compared to TVDI,the correlation coefficient between modified TVDI and soil moisture is 0.051 higher on average,and the standard deviation of soil moisture inversion is 0.13 lower.The modified method considering the vegetation coverage factor improves the accuracy of soil moisture inversion.