High Spectral Estimation Research Of Soil Salinity In Weifang Sea Salt Water Intrusion Area

Soil salinity seriously affected the effective use of land resources. Soil evaporation, geological features and hydro-climatic conditions and other natural factors and human and irrational use of common cause the formation of soil salinization. Distribution of vast, less land than the major are the salient features of the phenomenon of soil salinization in China, severely restricted the development of Chinese agriculture, soil salinity management is an important issue for the development of a harmonious society in China. Salt content of the soil is an important indicator of the level of soil salinization. Estimation of soil salt content can get real-time changes in soil salinity levels to prevent further deterioration of soil salinization.This article is in the National Natural Science Foundation, the national science and technology support program, Longkou mining area and perimeter monitoring by remote sensing in coastal zone research project, sea(salt) water intrusion area numerical modeling and forecast system development and construction projects and land resources remote sensing spectral library construction project in Shandong province, supported by Weifang sea study on estimation of soil salt for saltwater intrusion.Weifang sea salt water intrusion is located in the high efficiency ecological economy in Yellow River Delta area and blue one of the core of the economic zone of Northern District, Weifang city in Shandong peninsula, location is very important. Because of the low altitude, serious flooding with sea water, ground-water and high salinity and strong influence of factors such as evapotranspiration, soil salinization are obvious in the study area. Paper in field measured high spectrum data, and soil full salt content and high spectrum remote sensing data of support Xia, for has series of research: inquiry has different salinization degree Xia soil of spectrum features law, and on soil salt content for has features analysis, and on high spectrum measured data for has mathematics transform(spectrum features enhanced), and used single correlation analysis get has sensitive features band, and through multiple linear return analysis building has soil full salt content high spectrum estimates model, and established has soil spectrum library, and Information obtained with spectral angle classification of soil salinization. The main conclusions are as follows:(1) different levels of salinization of soil spectral characteristicsIn the 325-1075 nm within the band, different levels of salinization of soil spectral reflectance curve on the whole trend is consistent. As the wavelength increases soil spectral reflectance is getting bigger. Total salt content(that is, levels of salinization) higher and higher the spectral reflectance of each band. More back-band, reflectivity increases more obvious.(2) analysis of characteristics of soil salt contentsSoil physical and chemical analysis of samples showed higher levels of salinity in the study area. Salts of anions in soil dominated by Cl-and 46.15% of total anionic, cationic Na+-oriented, accounting for 64.85% of the total cation. Eight major ions and doing statistical analysis of the correlation of soil salt content, indicating that sea salt water intrusion in the study area within the Cl-in combination with Na+, Mg2+ and SO42-are the main way, salinity type is chloride-sulfate type.(3) this study hyperspectral data of the field conducted a series of transformations, spectral feature enhancements.On without transform of original spectrum reflection rate(r) for countdown(1/R), and logarithmic(lgR), and logarithmic countdown(1/lgR), and are square(?r), and package complex line elimination(CR) five species form transform, again on r, and 1/R, and 1/lgR, and?r and CR respectively for a order, and second-order differential processing, for next soil spectrum features of analysis and research do prepared.(4) the use of 18 kinds of spectra obtained in the previous step, simple correlation with soil salt content, access to the related diagrams and sensitive feature bands. Results show that a certain spectral transform can enhance hyperspectral data and correlation coefficient of soil salt content, which transforms the most obvious effect of the differential forms.(5) using multiple linear regression analysis of hyperspectral remote sensing estimation model of soil salt content. Results showed that the original square root second order differential reflectivity data processing data modeling best adjusted R2 of 0.9487.(6) through the measured hyperspectral data and field survey of soil spectral library, and hyperspectral remote sensing image data using spectral angular taxonomy on soil salinization information was extracted. Results show that the proper classification accuracy classification thresholds can be reached 73.2%.