Document Type : Research Article
Authors
isfahan university
Abstract
Now a day, performing soil surveys by minimum time and budget as well as enough accuracy is one of the soil scientists' interests. The present study was investigated to study the effect of sampling density on statistical and geostatistical accuracy of estimation for some of soil properties in Shahrekord plain. For this purpose, 240 soil samples (horizon A) with 125 m interval were collected. Then, the physical and chemical analysis included bulk density, soil texture, volume percent gravel, percentage of calcium carbonate equivalent, organic matter, pH and electrical conductivity were determined. The second (120 samples) and third (60 samples) sampling densities were randomly selected the first samplings (240 samples) by considering uniform distribution throughout the region. Results showed that there are no significant differences among the studied soil properties for three sampling intervals. Also, by reducing the number of samples, no regular increase or decrease was observed for coefficient of variation, mean estimation error and root mean square error. The cross validation showed that mean estimation error was close to zero and the root mean square error was low which indicate the acceptable accuracy of estimations. Visual and quantitative interpretation of kriging maps showed high accordance for three sampling densities. Therefore, semi-detailed soil surveys can be replaced for detailed ones which allow saving time and budget.
Keywords
)
Send comment about this article