Soil salinity characterization using hyper-spectral remote sensing data
KRISHI: Publication and Data Inventory Repository
View Archive InfoField | Value | |
Title |
Soil salinity characterization using hyper-spectral remote sensing data
Not Available |
|
Creator |
A. Barman
A.K. Mandal R. Srivastava R.K. Yadav P.C. Sharma |
|
Subject |
Soil Salinity
Hyper-spectral remote sensing |
|
Description |
Not Available
The complexity of salinization processes, spatial and temporal variability caused soil salinity mapping a difficult proposition. Severely salt affected soils (SAS) can be easily detected due to high reflectance from salt crust on soil surface, whereas, detection of low and medium SAS is difficult due to intricate association of salt, soil, water and vegetation. An attempt is made to characterize such SAS using hyper-spectral remote sensing data. A methodology was developed at CSSRI Karnal integrating hyper-spectral (HRS) data with limited ground truth and further quantifying through a statistical model. The variability of salinity and sodicity attributes such as ECe, Na+, Cl-, CO32- and HCO3- (me L-1) of the saturated soil extract were related quantitatively (r2=>90%) by hyper-spectral data. The spectral regions of 1400, 1900 and 2200 nm showed prominent peak due to the changes in soil salinity. At 1900 nm prominent shifting facilitated in establishing a significant correlation with salt concentration. A methodology was developed for characterizing salt affected soils using hyper-spectral data and is found useful for delineating salt affected soils right from the space platform (Fig. 1). (i) Study area (10.8 ha of CSSRI Experimental farm) is divided in regular grid (30 × 30 m) to collect surface soil samples and spectroradiometer data (ii) Soil samples were collected from the surface layers (0-30cm) during the post monsoon season and were analyzed for physico-chemical properties such as ECe (dS m-1), pHs, Na+, K+, Ca2+, Mg2+, CO32-, HCO3-, Cl- and SO42- (1) Hyper-spectral data were collected from a spectroradiometer in different wavelength regions and standardized using a statistical model to find prominent absorption region between 1420 to 2020 nm (2) Salinity model was developed integrating hyper-spectral data with soil physico-chemical properties by multivariate statistical data analysis and was validated using band math techniques (3) A spectral library thus developed for further mapping of salt affected soils with limited ground truth data. Not Available |
|
Date |
2018-02-09T10:14:18Z
2018-02-09T10:14:18Z 2017 |
|
Type |
News Letter
|
|
Identifier |
Barman, A., Mandal, A.K., Srivastava, R., Yadav, R.K. and Sharma, P.C. 2017. Soil salinity characterization using hyper-spectral remote sensing data. ICAR News 23(4): 11 – 12.
2394-3270 http://krishi.icar.gov.in/jspui/handle/123456789/5807 |
|
Language |
English
|
|
Relation |
Not Available;
|
|
Publisher |
ICAR NEWS
|
|