| Taking soil of seven-year-old forest of Eucalyptus dunnii Maiden in Jianyang City of FujianProvince as the research object. In the study area, make a rectangle area whose areage is660m*90m. In the north-south direction it is660meters long,and in the west-east direction it is90meters long. Cut up the rectangle area into many little squares whose side length is30m andcollect soil samples at each vertices of the square. In theory, there must be92soil sampling points,but with the restriction of the terrain and the distribution of Eucalyptus dunnii Maiden, there wereonly65sampling points. These points almost covered the growing area of seven-year-old forest ofEucalyptus dunnii Maiden. Collect soil samples in three different layers (0～20cm,20～40cm,40～60cm) at each point and measure their water content, electric conductivity and electricconductivity. Using descriptive statistical analysis method and Geostatistial method to study thespatial variation regulation of the soil water content, electric conductivity and electric conductivityin different scale. Then get the knowledge of the influence of the sampling scale to the spatialvariation regulation to choose the best sampling scale to interpolate. At the same time, comparethe Ordinary Kriging and Co-Kriging, and use the better one to interpolate. After the interpolation,draw pictures of the distributions of the soil water content, electric conductivity and electricconductivity. And the correlational analysis method was used to analyse the relationship amongwater content, electric conductivity and electric conductivity. All these study will do good for theplantation of Eucalyptus dunnii Maiden and the long-term maintenance of forest land productivity.The descriptive statistics analysis shows that there are some common ground and differencesin the soil water content, electric conductivity and electric conductivity in different samplingscales. The soil water content mean in different scale decreases with the increasing of soil depth,but there are obvious difference in the coefficient of variation and correlation of the soil watercontent in different layers.;The soil electric conductivity mean in different scale increases with theincreasing of soil depth, but there are obvious difference in the coefficient of variation andcorrelation of the soil electric conductivity in different layers; The soil electric conductivity meanin different scale decreases with the increasing of soil depth, and different sampling scale makesthe coefficient of variation change greatly. There are obvious difference in the coefficient of variation and correlation of the soil electric conductivity in different layers. By K-S nonparametrictest, the water content, electric conductivity and electric conductivity of different soil layers indifferent sampling scale show normal distribution. Comparing the best model of spatial variationand prediction errors of water content, electric conductivity and electric conductivity of differentsoil layers in different sampling scale, great difference appeared. According the rule of minimumerror, the best sampling scale (30m×30m) is chosen out for the next analysis. Using the dataunder30m×30m sampling scale to calculate the prediction errors of water content, electricconductivity and electric conductivity of different soil layers by Ordinary Kriging and Co-Krigingto choose a better method to interpolate. Then the result shows for the water content of0～20cmand20～40cm layers, Ordinary Kriging is better,while for40～60cm layer, Co-Kriging is better;for the electric conductivity of0～20cm layer, Co-Kriging is better,while for20～40cm and40～60cm layers, Ordinary Kriging is better; for the electric conductivity of0～20cm layer,Co-Kriging is better,while for20～40cm and40～60cm layers, Ordinary Kriging is better. Onbase of the above result, make a interpolation analysis and interpolation standard deviationanalysis, the result shows: the distribution of water content of0～20cm layer appears strippeddistribution,and its interpolation standard deviation distribution is increasing from the low valuecenter to its around, and the distribution of water content of20～40cm layer appears strippeddistribution along northwest-southeast direction with some high value and low value plaquesmixing in the stripped area, while the distribution of water content of40～60cm layer appearsstripped distribution without apparent direction; and its interpolation standard deviationdistribution is similar to the interpolation standard deviation distribution of20～40cm layer; thedistribution of electric conductivity of0～20cm layer is decreasing from the southwest to the eastand northeast, and the highest value appears in the southwest and the lowest value appears in thecenter, and for the distributions of electric conductivity of0～20cm and20～40cm layers, thehighest value appears in the southwest and northwest, and the electric conductivity is decreasingfrom west to east; the interpolation standard deviation of electric conductivity in different soillayers is small, and all appear the spatial distribution trends of taking the south and middle ascenter, and increasing gradually toward the northest. the distribution of electric conductivity of0～20cm layer is decreasing from the northwest to the southeast, and the highest value appears inthe northwest, and for the distributions of electric conductivity of20～40cm layer, the highest value appears in the southeast, and the electric conductivity is increasing from the lowest valuearea in center to southwest; the interpolation standard deviation distribution of electricconductivity in different soil layers is similar, and all appear the spatial distribution trends oftaking the south and middle as center, and increasing gradually toward the northest and thesouthwest.Comparing the best model of spatial variation and prediction errors of water content,electric conductivity and electric conductivity of different soil layers in different sampling scale,great difference appeared. According the rule of minimum error, the best sampling scale (30m×30m) is chosen out for the next analysis. Using the data under30m×30m sampling scale tocalculate the prediction errors of water content, electric conductivity and electric conductivity ofdifferent soil layers by Ordinary Kriging and Co-Kriging to choose a better method to interpolate.Then the result shows for the water content of0～20cm and20～40cm layers, Ordinary Krigingis better,while for40～60cm layer, Co-Kriging is better; for the electric conductivity of0～20cmlayer, Co-Kriging is better,while for20～40cm and40～60cm layers, Ordinary Kriging is better;for the electric conductivity of0～20cm layer, Co-Kriging is better,while for20～40cm and40～60cm layers, Ordinary Kriging is better. On base of the above result, make a interpolationanalysis and interpolation standard deviation analysis, the result shows: the distribution of watercontent of0～20cm layer appears stripped distribution,and its interpolation standard deviationdistribution is increasing from the low value center to its around, and the distribution of watercontent of20～40cm layer appears stripped distribution along northwest-southeast direction withsome high value and low value plaques mixing in the stripped area, while the distribution of watercontent of40～60cm layer appears stripped distribution without apparent direction; and itsinterpolation standard deviation distribution is similar to the interpolation standard deviationdistribution of20～40cm layer; the distribution of electric conductivity of0～20cm layer isdecreasing from the southwest to the east and northeast, and the highest value appears in thesouthwest and the lowest value appears in the center, and for the distributions of electricconductivity of0～20cm and20～40cm layers, the highest value appears in the southwest andnorthwest, and the electric conductivity is decreasing from west to east; the interpolation standarddeviation of electric conductivity in different soil layers is small, and all appear the spatialdistribution trends of taking the south and middle as center, and increasing gradually toward the northest. the distribution of electric conductivity of0～20cm layer is decreasing from thenorthwest to the southeast, and the highest value appears in the northwest, and for thedistributions of electric conductivity of20～40cm layer, the highest value appears in thesoutheast, and the electric conductivity is increasing from the lowest value area in center tosouthwest; the interpolation standard deviation distribution of electric conductivity in different soillayers is similar, and all appear the spatial distribution trends of taking the south and middle ascenter, and increasing gradually toward the northest and the southwest. |
PDF Full Text Request