AI Summary & Key Takeaways
AI InsightLinear, Gaussian and exponential models were used for different soil parameter based on nugget, mean prediction error and root mean square standardize prediction error values and maps were developed through extension techniques to cover all the area outside the sampling points.
📌 3 Key Takeaways
Abstract
Soil samples at 0-20 cm depth were collected from major crop areas of Hindukush mountainous range, District Chitral, extreme Northwestern Pakistan, during April 2014 to assess their physico-chemical properties and spatial distribution pattern. 103 soil samples were analyzed and maps were created by geostatistical technique of inverse distance weighting and kriging techniques using GIS and GS win-7 computer software. The soil texture ranged from silt loam to dominantly sandy loam, slightly acidic to alkaline and moderate to highly calcareous but with no salinity indication. Soil organic matter was higher than 2 % in about 75 % of samples. Soil pH, EC and lime showed slight dependence on each other with r values from 0.4 to 0.5 while OM varied independently as indicated by their lower correlation values. Semivariogram analysis showed that soil pH, lime, OM had strong spatial dependence (nugget-sill ratio, <25%) while silt, sand, EC had moderately (nugget-sill ratio, 25-75%) and clay had weakly distributed in the area. Linear, Gaussian and exponential models were used for different soil parameter based on nugget, mean prediction error and root mean square standardize prediction error values and maps were developed through extension techniques to cover all the area outside the sampling points.