Study On The Mechanism Of The Aggregation Scale Effect Of GlobeLand30 Spatial Data

Ground system consists of naturally occurring and complex geographical phenomena,Space and multi-scale representation of geographic phenomena make the description of spatial data a multi-scale features.Broadly speaking,spatial scale is the units to describe an object or phenomenon,it is the smallest length that can be identified in the image.As the scale of observation scale can not satisfy the production requirements,a rough estimation of error model is needed to achieve the conversion between two different scales.When the scale is translated in the ecological relationship between heterogeneous,information often will be distorted,it is how to introduce scale effect.There is a non-linear relationship and feedback mechanisms between different scales,including the spatial heterogeneity,the diversity of the scale,and morphological diversity of patches,they make the surface structure seem regular in the macro,and heterogeneity in the micro perspective.The study on Mechanism of scale effect is helpful to grasp the scale problem and reveal the complex and diverse variation of surface features,and establish the error estimation model.At present,the scale effect in qualitative research is more,quantitative research is less,the article will reveal the mechanism of scale effects through composition and structure of the surface morphology,and quantify and analyze of multi-scale effects scale in the geometrical morphology and ecology discipline,and analyze of the uncertainty of multi-scale representation finally.In this paper,GlobeLand30 Spatial Data is Used for this study,the feature-rich information is choose as sample,through the analysis of the shape index of different patches,this paper establish morphological classification model,the image and feature database is as the base data,then feature extraction is completed.In this paper,the patch is divided into point data,large line data,small line data,Compact convex polygon,general polygon and irregular polygon,and analyzed different types of scale-dependent characteristics,through establishing the interpolation regression model for each category under different scale to reveal the different categories scaling law.The method achieve the research of micro-data upscaling from the perspective of single factor,namely it removes the mutual restricted relationship between data,the spatial characteristics seem as a homogeneous.On this basis,the upscaling performance of the various elements in a single category is fitted,and then we get fitting results and fitting error of every single category.Then it analyzes the neighborhood relations between the individual elements,by studying the micro and mid data to fit the upscaling rules in macro,at last,experimenting results and verifying the fitting error is a must.Experiments show that the fitting error of the area ratio' standard deviation about nodata,arable land,water,and artificial cover is 0.083,0.008,0.062,0.013,it is the change information that all categories of the experimentation area deviated from homogeneous spatial structure in upscaling process,that is the deviation from scale effect.Finally,the paper studies the uncertainty of multi-scale representation.Due to the uncertainty of multi-scale representation is made form of nonlinear data processing model and the complexity of geographic systems,experiments show that uncertainty from scale effect is more significant.The degree of deviation about information expression from homogeneous space is studied in this paper through micro,mid,macro view,what makes up the defects of pure research from the macroscopic angle.To quantify the scale effect Reveal the mechanism of Scale effect,and provides a new way for the study on the scale effect.