基于地理探测器和GIS的压砂地砾石空间异质性及其影响因素分析

Determining the spatial heterogeneity of gravel and its influencing factors in a gravel-sand mulched field using geographical detector and GIS

为探究压砂地不同粒径砾石空间异质性及其影响因素,基于压砂地102个采样点,对不同粒径砾石配比进行了描述性统计,通过半方差函数和全局空间自相关分析等地统计方法分析了压砂地不同粒径砾石的空间变异特征和聚集模式,并进一步利用地理探测器模型识别影响不同粒径砾石占比空间分异的主要影响因子,在此基础之上,选取主要影响因子协同地理加权回归克里格法对不同粒径砾石占比进行可视化,获取其分布特征。结果表明,不同粒径砾石的空间变异系数均大于10%,属中等变异。不同粒径砾石的空间异质比变化范围较大(0.26%~83.48%),其空间异质性受随机性因素和结构性因素共同影响。空间自相关分析显示各粒径砾石的全局Moran′s I均高于0.674,且Z值均大于1.96,表明不同粒径砾石具有极显著的空间依赖特征,呈聚集性分布。地理探测器模型识别表明压砂地不同粒径砾石的空间分异的解释力由大到小依次为种植年限>坡向>地表粗糙度>坡度>剖面曲率>平面曲率>地表起伏度,种植年限的单因子解释力明显高于其他因子。不同环境条件下压砂地土体中不同粒径砾石空间分布差异明显且呈聚集性分布特征。种植年限、坡向和地表粗糙度是压砂地不同粒径砾石空间分布状况的主要影响因子。

This study investigated the spatial heterogeneity of gravel with different particle sizes and the factors influencing these in gravelsand mulched fields. Based on 102 sampling points in a gravel-sand mulched field, descriptive statistics of gravel with different particle size were conducted to analyze the characteristics of the spatial variation and aggregation patterns of gravel with different particle sizes via two geostatistical methods, semi-covariance function and global spatial autocorrelation analysis. Then, the main factors influencing the spatial variation of gravel proportion were identified using a geographic probe model. The main influencing factors were then used to visualize the percentage of gravel with different particle sizes and obtain their distribution characteristics via cooperative geographically weighted regression kriging. The results showed that the spatial variation coefficients of gravel with different particle sizes were all greater than 10%, which was a moderate variation. The spatial heterogeneity ratios of the gravel size varied over a wide range(0.26%–83.48%),indicating that their spatial heterogeneity and variability were influenced by both stochastic and structural factors. Spatial autocorrelation analysis showed that the global Moran′s I index of gravel of each particte size was higher than 0.674 and the Z-values were all greater than1.96, indicating highly significant spatial dependence characteristics for gravel of different particle sizes with aggregated distribution. The explanatory power of the spatial variance of different particle sizes in the gravel-sand mulched fields was in the order of planting year>slope aspect> surface roughness> slope> profile curvature> plane curvature> relief degree of land surface. The single factor explanatory power of planting year was significantly higher than that of the other factors. The spatial distribution of gravel with different particle sizes in the soil of the gravel-mulched field varied significantly under different environmental conditions, with the distribution appearing aggregated. The planting year, slope aspect, and surface roughness were the main factors influencing the spatial distribution of gravels of different particle sizes in gravel-sand mulched fields.