The Inversion And Analysis Of Soil Moisture Content In Coal Mining Area Based On Nir-Red Spectral Feature Space

As an important link of surface water circulation,soil moisture plays a significant role in maintaining a healthy development of regional ecological environment.The distribution of soil moisture is mainly affected by two aspects,i.e.natural conditions and human activities,especially,the changes in soil moisture caused by coal mining cannot be ignored.The ground deformation and changes of stratum structure caused by coal mining can change the movement of soil moisture,and then affect the growth conditions of vegetation and even destroy local ecological environment.The Daliuta mining area has a wide range of coal mining activities throughout the year,due to little precipitation and dry surface,the ecological environment of this region is quite fragile,and coal mining activities further intensify the drought condition and deteriorate eco-environment.Therefore,it is particularly important to study the effects of coal mining on the loss of soil moisture and ecological environment in study area by analyzing the temporal and spatial changes characteristics of soil moisture content.With abundant available data sources and lower costs,optical remote sensing becomes an efficient way to monitor soil moisture in large scale.This paper selected the Daliuta mining area located in the Shenmu county of Shaanxi province as the study area,employed the Landsat remote sensing data and in situ soil moisture content and spectral data to study inversion methods of soil moisture in the arid mining area with little covered vegetation.On the basis of analyzing the sensitive bands to soil moisture,developing the Nir-Red spectral feature space and comparing the drought index models based on the feature space,the Angle Dryness Index Model was improved and validated through measured soil moisture and spectral data in the study area.According to the validation results and practical situation of the study area,the optimal model was determined to retrieve multi-temporal soil moisture,by studying the temporal and spatial characteristics of soil moisture and the relationships between soil moisture and vegetation,the effect of mining activities on the soil moisture and vegetation were analyzed.The main results of this paper are as following:(1)The correlation relationship of measured soil moisture and spectral data in study area showed that there existed significant correlation between soil moisture content and remote sensing reflectance at Red band and Nir band with R of-0.7987 and-0.8029,respectively,which are the sensitive bands to soil moisture.(2)The nonlinear model(ADI)was improved by quadratic polynomial model(MADI)that was developed based on fitting equations between angle index(?)and soil moisture content in the Nir-Red spectral feature space,and the validation results indicated that MADI had a higher inversion precision.(3)Five kinds of dryness index model based on Nir-Red spectral feature space were established and validated through in situ data,and the validation results showed that the MADI model had the wider applicability and higher inversion precision,and it was determined as the most appropriate soil moisture retrieval model in the study area.(4)The results showed that the soil moisture content levels in the study area gradually decreased from 2002 to 2015,and the decrease range from 2009 to 2015 was lower than that of from 2002 to 2009.The change of soil moisture showed a corresponding relationship to the change of covered vegetation in the whole,which indicated that the mining activities can cause the decrease of soil moisture and vegetation coverage.