eduzhai > Applied Sciences > Computer Science >

Prediction of Soil Salinity Using Multivariate Statistical Techniques and Remote Sensing Tools

  • carsar
  • (0) Download
  • 20210407
  • Save

... pages left unread,continue reading

Document pages: 14 pages

Abstract: Soil salinity limitsplant growth, reduces crop productivity and degrades soil. Multispectral datafrom Landsat TM are used to study saline soils in southern Tunisia. This studywill explore the potential multivariate statisticalanalysis, such as principal component analysis (PCA) and cluster analysis to identify the most correlated spectral indices and rapidly predict saltaffected soils. Sixty six soil samples were collected for ground truth data inthe investigated region. A high correlation was found between electricalconductivity and the spectral indices from near infrared and short-waveinfrared spectrum. Different spectral indiceswere used from spectral bands of Landsat data. Statistical correlation betweenground measurements of Electrical Conductivity (EC), spectral indices andLandsat original bands showed that the near and short-wave infrared bands (band4, band 5 and 7) and the salinity indices (SI 5 and SI 9) have the highestcorrelation with EC. The use of CA revealed a strong correlation betweenelectrical conductivity EC and spectral indices such abs4, abs5, abs7 and si5.The principal components analysis is conducted by incorporating the reflectancebands and spectral salinity indices from the remotesensing data. The first principal component has large positive associations withbands from the visible domain and salinity indices derived from these bands,while second principal component is strongly correlated with spectral indicesfrom NIR and SWIR. Overall, it was found that the electrical conductivity EC ishighly correlated (R2 = -0.72) to the second principal component (PC2), but no correlation is observed between EC and the first principalcomponent (PC1). This suggests that the second component can be used as anexplanatory variable for predicting EC. Based on these results and combining the spectral indices (PC2 and absB4) into a regression analysis, model yielded arelatively high coefficient of determination R2 = 0.62 and a lowRMSE = 1.86 dS m.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×