|
Investigating the efficiency of Geostatistical methods for monitoring and zoning some chemical properties of soil
|
Noshin Shahinzadeh   , Teimour Babaeinejad , Kamran Mohsenifar * 1, Navid Ghanavati |
|
|
|
Abstract: (829 Views) |
Soil chemical properties variability and value of their variability have specific importance in sustainable production. Measuring methods of these which usually possible with soil testing, in a large scale are so timing and cost consuming. One of methods which use for monitoring these properties and decrease the costs of sampling in large scale is interpolation. Interpolation has many methods and each one has a different accurate in various conditions. In current research three ordinary interpolation methods were used for interpolation some chemical properties of Khuzestan’s agricultural soils. These variables in this research were Total Nitrogen (TN), Absorbable Phosphorous (P), Absorbable Potassium (AK), Calcium Carbonate percent (CaCo3%), Organic Carbon percent (OC%), Electrical Conductivity (EC), and Soil Acidity (pH), which measured in 270 soil samples.
Due to the skewness and Kurtosis of these data, it was found they have not a normal distribution. In order to normalize the data, after deleting the outdated data, logarithm and Cox-box method were used.
The interpolation methods which used were kriging (OK), Radial Basic Functions (RBF), and Inverse Distance Weighted (IDW). Mean absolute error (MAE), mean bias error (MBE), and root mean square error (RMSE) were used for detecting the optimum method.
Examination of variographic results showed that the spatial structure (correlation ratio) of absorbable phosphorus (0.243) and absorbable potassium (0.159), strong, total nitrogen weak (measurement error 0.816) and the other parameters had a moderate spatial structure. Based on the results of this table, the ordinary kriging method with exponential model can be considered as the best and most accurate method (MAE = 0.01515, RMSE = 0.0890, MBE = 0.0006 and R2 = 0.912) for interpolation the total nitrogen.However, the best method for interpolation of organic carbon, absorbable phosphorus, absorbable potassium, calcium carbonate, acidity and electrical conductivity of soil are RBF, ordinary kriging (spherical model), ordinary kriging (spherical model), RBF, ordinary kriging (with exponential model) and RBF respectively.The results showed that the IDW method was not a suitable method for interpolating any of the parameters.
From the prepared maps, it can be clearly seen the non-uniformity of fertilization, amount of absorbable phosphorus in the southern parts of the study area is higher but amount of absorbable potassium in the northern and central parts of the study area has been more.While in the eastern and western part of the study area and in places where the pH is above 7.7, the amount of calcium carbonate in soils is at its highest level of 46 to 49 percent.
Preparing maps with appropriate accuracy by geostatistical methods for parameters that have good spatial dependence is a suitable method for monitoring and land management at the macro level. |
|
| Keywords: Spatial variability, zoning, soil chemical properties, interpolation, geostatistics |
|
|
Full-Text [PDF 1103 kb]
(196 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2021/05/26 | Accepted: 2021/10/26 | Published: 2023/01/30
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
