An integration of soil characteristics by using GIS based Geostatistics and multivariate statistics analysis Sultan Batheri block, Wayanad District, India

Abstract

The study was to estimate the spatial variability of physico-chemical characteristics of soil by using GIS-based geostatistics and multivariate analysis in Sultan Batheri Block, Wayanad district. A combination of geostatistics and multivariate statistical analysis was used to scrutinise and compare the soil data for spatial assessment investigation. In total, 32 soil samples were collected during the summer season (March to May 2021), from different locations through random sampling. Analyzing the characteristics was determined using various standard analytical techniques. The multivariate statistical analysis investigated the excellent results of soil properties. The principal component analysis and correlation coefficient matrix analysis determined the interrelationships of soil properties and showed a good variance of SOC. The highest variation was observed in the SOC whereas the lowest variation was observed in Electrical Conductivity. In PCA analysis, we found that the KMO value is exactly following around 0.788 or slightly below, which is also not exactly optimal. However, since the value is higher than 0.5, soil data factor analysis might be helpful. There are two statistical analyses were used to depict the soil properties and exemplify the spatial variability of soil characteristics in a GIS environment. Inverse distance weighting and kriging were used to develop the spatial distribution of all soil characteristics. According to a prediction generated using ordinary kriging, the SOM shows a high level of spatial autocorrelation in a geostatistical analysis with an elevation range of 2261–2661 m. The R2 coefficient for all soil properties ranges from 0.0072 to 0.2862. The soil variable SOM and the elevation range of 2224 m have a significant spatial autocorrelation in the Kriging semivarigram model. The R2 coefficients for all soil properties ranged from 0.0072 to 0.3162. The interpolation method was relatively accurate for all soil parameters in both types. The cross-validation test results showed that the prediction of soil characteristics was of high accuracy. The levels of nitrogen and phosphorus in the study region are ideal. The study ensured that nitrogen, phosphorus, and other insecticides and herbicides were not used. The result reveals that there is no risk of contamination from nitrogen and phosphorus and that soil is unquestionably contamination-free.