Soil Salinity Content In Weibei Area Based On Field Measured Spectral Monitoring

Soil resources are the important natural resources,which can't be replaced.Human life,social economic growth,the development of regional agriculture is affected by the quality and quantity of soil.Human unreasonable behavior lead to soil problems appear constantly,the phenomenon of soil salinization in coastal area affected by the local people's production and living,caused great damage to the ecological environment.The observational studies of soil salt content have important practical significance in coastal areas,which can get the space distribution of salt content and temporal variation of salt effective information.In this paper,the study area is the north of Wei-fang area,Under the support of the National Natural Science Fund Project(41371395),Shandong Province Geological Environmental Monitoring Station “the Yellow River delta Efficient Ecological Economic Zone sea salt water intrusion investigation and monitoring and early warning system construction”,Shandong Province Science Fund “Construction project of remote sensing spectral database of land and resources in Shandong Province” and “Research on remote sensing monitoring of Longkou mining area and the surrounding coastal zone”.The models of Soil salt parameters by multiple linear regression and principal components analysis were established by spectral data and laboratory test data,combined with different forms of spectral index;The optimal models were used to inversion calculate the salt parameter;Salt parameter spatial distribution characteristics are analyzed using statistical methods,parameter distribution was made with the aid of kriging interpolation method.The main research contents and research results are as follows:(1)111 soil samples were set in the north of Wei-fang,53 soil samples measured spectral data,soil salt content been assayed at the same time.Considering the project funds and other reasons,remaining 58 samples are estimated samples,only spectral data,not the physical and chemical experiments.(2)According to the physical and chemical properties of 35 soil samples and spectral data transformation,the models of multiple linear regression and principal components analysis were obtained,and choose the optimal estimation model and method.Results show that there is the best effect according to the second order differential transform index.The decisive coefficients of stepwise regression model about the salt content,K+,Na+,Ca2+,Mg2+,PH were 0.984,0.916,0.963,0.791,0.976,0.901 using second-order differential transform indicators.The decisive coefficients of the test samples were 0.7208,0.7177,0.7331,0.5845,0.6003 and 0.644.The decisive coefficients of principal component model about Cl-,SO42-and HCO3-were 0.869,0.821 and 0.601 using second-order differential transform indicators.The decisive coefficients of the test samples were 0.7654,0.5923 and 0.7204.Results show that selecting the modeling methods and modeling index in the study area has high universality and can provide theory for the research on quantitative monitoring of soil salt in the study area.(3)Soil salt contents were inversion calculated Combined with the spectral data of 58 evaluate samples using the optimal models.The average value of the salt content,K+,Na+,Ca2+,Mg2+,Cl-,SO42-,HCO3-,PH were 1209,19.983,260.751,75.208,45.763,314.42,300.51,273.63,7.337.(4)The results of descriptive statistics about the salt content,K+,Na+,Ca2+,Mg2+,Cl-,SO42-,HCO3-,PH showed that: The coefficients of variation(CV)varied from 2.99% for PH to 314.42%for Cl-and according to the coefficients of variation(CV)classification,PH had lowvariability(CV<10%),whereas total salt,Cl-,K+,Mg2+,Na+ and SO42-had high variability(CV >100%).Ca2+ and HCO3-had the moderate spatial variability(10%? CV ? 100%).(5)The variogram models for different variables of soil were obtained by using the Geostatistical software(GS + 9.0).The results of the spatial structural analysis showed the optimal model,with Spherical model for total salt,Exponential model for Cl-,Na+,SO42-,and Gaussian model for Ca2+,HCO3-,K+,Mg2+,PH.Total salt,Cl-,K+,Mg2+,Na+,PH had moderate spatial dependence;Ca2+,HCO3-,SO42-had strong spatial dependence.(6)The results of spatial interpolation show that: a decreasing trend existed from the northeastern to southwest,suggesting that total salt may be attributed to seawater,and the farther the distance from the sea,the lower soil total salt.The maps of K+,Na+,Cl-,Mg2+,SO42-and HCO3-displayed quite similar trend with a gradually increases from inland plain to coastal plain.SO42-and HCO3-showed an increasing trend,but the trend was not clear.The spatial distribution of PH and Ca2+ were displayed the shape of patch.