Accuracy analysis of remote sensing index enhancement for SVM salt inversion model
Authors:
- 15 November 2021
- journal article
- research article
- Published by Taylor & Francis in Geocarto International
- Vol. 37 (8) , 2406-2423
Abstract
The accurate and rapid acquisition of the soil salt content and spatial distribution is of great significance for effectively monitoring cropland soil salinization. Using OLI/Landsat-8 images as data sources, shortwave infrared bands (SWIR1, SWIR2) are introduced to the traditional vegetation and salinity indices, and the enhanced indices are put forward to obtain an improved inversion effect. After enhancement, the correlation coefficient between the indices and soil salt content increased by 0.01-0.44, and the multicollinearity between indices significantly decreased. The prediction ability of the support vector machine salt inversion model improved from general to fairly good, and the predicted and measured values were close to a linear distribution. The predicted and measured values ranged from 0.39-36.38 g·kg−1 and 0.35-38.09 g·kg−1, respectively, and 63.25% of the regional salt status could be accurately predicted. The results showed that the accuracy of soil salinity inversion can be effectively enhanced by introducing shortwave infrared bands.Keywords
Funding Information
- Research and Demonstration of Fertilization Regulation and Soil Conditioning Technology in Greenhouse Cucumber (19226912D)
- Hebei Talent Engineering Training Support Project (A201910003)
This publication has 43 references indexed in Scilit:
- Monitoring and evaluating spatial variability of soil salinity in dry and wet seasons in the Werigan–Kuqa Oasis, China, using remote sensing and electromagnetic induction instrumentsGeoderma, 2014
- Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries: Applications in a date palm dominated regionGeoderma, 2014
- Regional land salinization assessment and simulation through cellular automaton-Markov modeling and spatial pattern analysisScience of The Total Environment, 2012
- Comparison Between Active Learning Method and Support Vector Machine for Runoff ModelingJournal of Hydrology and Hydromechanics, 2012
- Predicting soil chemical composition and other soil parameters from field observations using a neural networkComputers and Electronics in Agriculture, 2012
- The use of remote sensing in soil and terrain mapping — A reviewGeoderma, 2011
- Regional‐scale Assessment of Soil Salinity in the Red River Valley Using Multi‐year MODIS EVI and NDVIJournal of Environmental Quality, 2010
- Determining the composition of mineral-organic mixes using UV–vis–NIR diffuse reflectance spectroscopyGeoderma, 2006
- Detecting salinity hazards within a semiarid context by means of combining soil and remote-sensing dataGeoderma, 2006
- Image-derived spectral endmembers as indicators of salinisationInternational Journal of Remote Sensing, 2003