|
Soil salinization led to various negative effects such as regional environmental quality change and ecological balance and so on, and also caused a series of natural disasters and the grain issues, etc, which maed against sustainable development of social economy, so the study on soil salinization become an pressing important problem to be resolved how to rationally develop and utilize salinatefields. This paper taked the lower Reaches of Kaidu River Basin as study area, the research object is salinized soil.Adopt RS and GIS technology, combined with field survey, collected samples in representative areas, used indoor experimental analysis, analyzed total salt content of the soil samples, the main ions (HCO3-,CL-,SO42-,Ca2+,Mg2+和Na++K+) , PH , TDS, electric conductivity indicators by multivariate statistical methods. The salinatefields’spectral information of 1999 2009 Landsat ETM + images was extracted by supervised classification. Quantitatively analyzed spectral information of space and time scales dynamic changes of inter annual salinization in recent 12 years.Through the statistical software SPSS16.0 and DPS, analyzed the relation between remote sensing image the spectrum characteristics value and salt contents by principal component analysis and correlation analysis. Collected soil samples with different types and degree by ASD high spectrometer, and analyzed the high spectral characteristics of those soil samples.Through analyzing, six aspects of the results were concluded as follows:（1）In study area, CL--SO42- saline soil is mainly in study area, which accounts for 43.75 %, the following is SO42- saline soil ,which account for 35.98 %.（2）Statistical analysis shows that, the correlation coefficients between SO42- Ca2+and Mg2 +,Na++ K+ are 0.909,0.904,0.867 respectively in the soil surface, the correlation coefficient between CL-and Na++K+ is 0.957 presents significant positive. In the soil surface, sulphide and chloride accumulate intensely, salt move up primarily.（3） Established the linear relationship between spectral characteristic value of the remote sensing images and soil salinity: y = 13.91x + 62.38(R2≥0.719.（4）CL--SO42- saline soil and SO42- saline soil shows different high spectral characteristics with different levels.Reflectance curve characteristic points of different levels CL--SO42- saline soil: five peaks （located in 550nm、1050nm、1175nm、1550nm、2125nm）and four valley values （located in980nm、1100nm、1400nm、1900nm）. Along with the increase in degree salinization, the reaction in nearly is obvious at 550nm band, the difference of the spectrum curve eigenvalues is minimal at 1400nm2500nm. Reflectance curve characteristic points of different levels SO42- saline soil: five peaks （located in 550nm, 750nm ,800nm, 1150nm, 1250nm）and five valley values （located in 775nm, 975nm, 1125nm, 1400nm）.With varying degrees of sulfuric acid of SO42- saline soil, the reaction in near-infrared wavelength is obvious,the difference of four valley values corresponding to 1100nm, 1400nm, 980nm, 1900nm is minimal.（5） 19992009, the area of severe saline land 351.75 km2 had reduced to 194.44 km2 , the reducing area was 157.31km2, the average annual rate of change was 4.07%; The area of moderate saline had increased to383.02km2 from 264.88 km2, the area of increasing is 118.14 km2, the average annual rate of change was 4.05%; the area of mild saline soil has increased to 761.59km2 from 321.57km2 km2, the area of increasing is 440.02 km2, the average annual rate of change was 12.44%.（6） The interannual transformation between non-saline soil and mild saline soil is frequent and fierce, the transformation percentage is more than 16%.
|