Application of a geographically-weighted regression analysis to estimate net primary production of Chinese forest ecosystems

Aim The objective of this paper is to obtain a net primary production (NPP) regression model based on the geographically weighted regression (GWR) method, which includes spatial non‐stationarity in the parameters estimated for forest ecosystems in China. Location We used data across China. Methods We examine the relationships between NPP of Chinese forest ecosystems and environmental variables, specifically altitude, temperature, precipitation and time‐integrated normalized difference vegetation index (TINDVI) based on the ordinary least squares (OLS) regression, the spatial lag model and GWR methods. Results The GWR method made significantly better predictions of NPP in simulations than did OLS, as indicated both by corrected Akaike Information Criterion (AIC_c) and R². GWR provided a value of 4891 for AIC_c and 0.66 for R², compared with 5036 and 0.58, respectively, by OLS. GWR has the potential to reveal local patterns in the spatial distribution of a parameter, which would be ignored by the OLS approach. Furthermore, OLS may provide a false general relationship between spatially non‐stationary variables. Spatial autocorrelation violates a basic assumption of the OLS method. The spatial lag model with the consideration of spatial autocorrelation had improved performance in the NPP simulation as compared with OLS (5001 for AIC_c and 0.60 for R²), but it was still not as good as that via the GWR method. Moreover, statistically significant positive spatial autocorrelation remained in the NPP residuals with the spatial lag model at small spatial scales, while no positive spatial autocorrelation across spatial scales can be found in the GWR residuals. Conclusions We conclude that the regression analysis for Chinese forest NPP with respect to environmental factors and based alternatively on OLS, the spatial lag model, and GWR methods indicated that there was a significant improvement in model performance of GWR over OLS and the spatial lag model.     

基于地理加权回归模型的武汉城市圈生态用地时空演变及影响因素

生态用地对城市圈可持续发展至关重要。本研究以武汉城市圈32个研究单元生态用地为对象,采用土地利用转移矩阵、探索性回归分析和地理加权回归模型(GWR),首先利用遥感影像解译成果对2000—2005、2005—2010和2010—2015年各个单元生态用地的时空演变进行统计,然后利用公司企业、生活服务等点位及数量大数据完善传统影响因素指标体系,并进行探索性回归分析,精选最优回归模型,最后基于GWR模型对不同时期影响因素及空间分异规律进行分析。结果表明: 2000—2015年,城市圈内生态用地非生态转化呈现先升后降的倒“U”形变化规律,空间上呈现由点到面的扩张趋势;城市圈内共有8.4%的土地利用类型发生了转化,其中,耕地、林地、草地、水体和未利用地向非生态用地转型量占41.9%;空间格局由武汉中心城区逐渐向市级次中心、县级城镇周边扩展。探索性回归分析3期的通过模型数为326个,对所有模型进行GWR和普通最小二乘法(OLS)回归比较分析,3期最优模型的调整R²分别为0.83、0.91和0.76,前者较后者提高了0.02、0.03和0.02,AICc值分别减小2.88、3.42和0.83。GWR模型结果表明,武汉城市圈内生态用地转化影响因素的空间分异明显,影响模式在空间上以不同方向的逐渐过渡为主,兼有“V”形分布等其他模式。空间因素影响效果显著,空间数据潜在信息增强了城市圈内生态用地演化的解释力度。     

Using water and energy variation to explain the botanical richness pattern of Theaceae species in southern China

BackgroundEnergy and water availability are essential for biodiversity maintenance. In addition to the independent effects of water and energy on biodiversity, recent studies clarified that the effects of interaction between water and energy availability were indispensable.MethodsIn this exercise, by combining the species presence information and the environmental predictors, we produced species distribution models at 20 × 20 arc-minute resolution for 193 Theaceae species. Initially, the ordinary least square (OLS) regression was used to examine the stationary relationships between Theaceae diversity and climate. The statistical effects of water and energy on species diversity were detected using Geographically Weighted Regression analysis (GWR). Furthermore, the contour plots were used to view the statistical effects of the water and energy interaction on species diversity.ResultsThe OLS results suggested that both energy and water availability are related to Theaceae species diversity. In GWR regression, the spatial variation of energy and water showed high explanatory power to the diversity pattern of Theaceae species. The patterns in the residuals of both OLS and GWR regression varied geographically. Therefore, the results of GWR regression were kept for further analysis. The value of diversity-water slopes decrease changed from positive to negative in extremely wet regions; In extremely dry conditions, the value of diversity-energy slopes decrease faster than other regions.ConclusionsOur results support the following findings: 1) the latitudinal distribution of Theaceae species was limited by thermal tolerance, which support the freezing-tolerance hypothesis in macro-ecology; 2) Theaceae species diversity are sensitive to the instability of precipitation, while the limitation from energy availability is weak; 3) the effects of water and energy on species diversity are strong in dry regions. Those findings can provide further implications for Theaceae species conservation under climate change scenarios.     

基于地理加权回归克里格模型的帽儿山地区森林碳储量空间分布

森林碳储量对于全球气候变化具有重要影响,以往的模型估算未考虑到模型残差的空间相关性和碳储量数据的非平稳性,影响模型的预测精度.本研究基于东北林业大学帽儿山实验林场的ETM+遥感影像数据和193块固定样地,利用地理加权克里格回归(GWRK)建立森林碳储量与遥感和地形因子的回归模型,同时对比最小二乘模型(OLS)、地理加权回归模型(GWR)的预测精度.结果表明: 对于帽儿山地区的森林碳储量估算,GWRK的平均绝对误差(MAE)、均方根误差(RMSE)低于OLS模型和GWR模型,GWRK模型的平均误差(ME)低于GWR模型,与OLS模型相近.GWRK模型的预测精度为83.2%,较OLS模型(73.7%)和GWR模型(77.3%)分别提高6%和10%,拟合精度明显提高,说明GWRK模型是森林碳储量估算的有效方法.利用GWRK模型预测的研究区森林碳储量平均值为70.31 t·hm^-2,在海拔较高的地区,森林碳储量值相对较高,说明海拔对其有较大影响.     

Extending geographically and temporally weighted regression to account for both spatiotemporal heterogeneity and seasonal variations in coastal seas

Space-time modelling has been successfully applied in numerous research projects and has been studied extensively in the field of geographical information science. However, the cyclical or seasonal variations in the temporal dimension of most spatiotemporal processes are rarely considered along with spatiotemporal nonstationarity. Seasonal variations are widespread and typical in marine environmental processes, and addressing both spatiotemporal heterogeneity and seasonal variations is particularly difficult in the turbid and optically complex coastal seas. By incorporating seasonal periodic effects into a geographically and temporally weighted regression (GTWR) model, we proposed a geographically and cycle-temporally weighted regression (GcTWR) model. To test its performance, modelling of chlorophyll-a, known as an important indicator of the coastal environment, is performed using the in situ data collected from 2012 to 2016 in the coastal sea of Zhejiang Province, China. GcTWR is compared with global ordinary least squares (OLS), geographically weighted regression (GWR), cycle-temporally weighted regression (cTWR), and GTWR models. In the results, the GcTWR model decreases absolute errors by 89.74%, 79.77%, 76.60% and 29.83% relative to the OLS, GWR, cTWR, and GTWR models, and presents a higher R² (0.9274) than the GWR (0.5911), cTWR (0.6465), and GTWR (0.8721) models. The estimation results further confirm that the seasonal influences in coastal areas are much more significant than the interannual effects, which accordingly demonstrates that extending the GTWR model to handle both spatiotemporal heterogeneity and seasonal variations are meaningful. In addition, a novel 3D visualization method is proposed to explore the spatiotemporal heterogeneity of the estimation results.     

Spatial pattern of non-stationarity and scale-dependent relationships between NDVI and climatic factors—A case study in Qinghai-Tibet Plateau,China

Spatial non-stationarity and scale-dependence are important characteristics of the relationship between NDVI and climatic factors. To improve the reliability of model prediction, it is necessary to find the scales and spatial heterogeneity in which a stationary relationship is reached. In this paper, a geographically weighted regression (GWR) model was developed to define spatial non-stationarity and scale-dependent relationships between NDVI and climatic factors. The results indicate that the spatial scale of the stationary relationship for NDVI and both temperature and precipitation is 156 km over the whole Qinghai-Tibet Plateau. Both modeling performance and the spatial pattern of the GWR model are significantly better than global regression models such as OLS. Significant spatial heterogeneity of regression relationships between NDVI and climatic factors is revealed within the Qinghai-Tibet Plateau. We conclude that the dominant climatic factor influencing NDVI is not the same for all ecoregions within the study area. There are also different key scales of interaction between NDVI and the dominant climatic factor in these various ecoregions. Finally, model performance is different in the each eco-region. Therefore, this finding can provide a scientific basis for choosing a suitable scale and reliable models to solve scale-dependent problems in geography and ecology.     

基于地理加权回归模型探究环境异质性对秦岭大熊猫空间利用的影响

环境异质性对野生动物分布的影响具有明显的空间不均匀性。传统分析中多采用经典线性回归模型来量化野生动物分布与环境变量之间的关系,难以准确反映物种-环境关系的空间异质特征。地理加权回归(GWR)是近年来提出的一种新的空间分析方法,通过将空间结构嵌入线性回归模型中,以此来探测空间关系的非均匀性。以秦岭大熊猫为例,应用GWR模型分析大熊猫空间分布与环境异质性特征之间的潜在关系,并同经典的全局最小二乘回归法(OLS)进行比较。结果表明,GWR模型的AIC、R²和校正R²均显著优于OLS模型,GWR模型的局部回归系数估计能够更加深刻地揭示大熊猫空间分布与环境变量间的复杂空间关系,且GWR模型能够为物种的科学保护提供更加有效的理论支撑。因此,GWR模型可为探究物种-环境关系的空间异质特征提供一种新的方法,在物种栖息地选择与利用研究中具有一定的应用前景。     

基于不同空间模型的马尾松林生态系统碳密度研究

以江西省马尾松林生态系统为研究对象,基于样地调查及样品碳含量测定结果计算其碳密度,并选取立地、植被及气象等方面的15个因子,采用多元线性逐步回归方法筛选出对生态系统碳密度影响显著的因子,然后分别利用最小二乘模型(OLS)、空间误差模型(SEM)、空间滞后模型(SLM)和地理加权回归模型(GWR)构建生态系统碳密度与其影响因子之间的关系模型,筛选出最优的拟合模型。结果表明:对马尾松林生态系统碳密度影响显著的因子分别为海拔、坡度、土层厚度、胸径、年均温度和年均降水量。4种模型拟合结果均显示碳密度与坡度呈负相关,与海拔、土层厚度、胸径呈正相关。模型的决定系数(R²)由大到小分别为GWR(0.8043)>SEM(0.6371)>SLM(0.6364)>OLS(0.6321),模型均方误差(MSE)与赤池信息准则(AIC)最大的均为OLS模型,最小的均为GWR模型;残差检验表明GWR模型能有效降低模型残差的空间自相关性。综合分析得出GWR模型的拟合效果最优,更适用于江西省马尾松林生态系统碳密度的估测。     

Geographic Disparity in Chronic Obstructive Pulmonary Disease (COPD) Mortality Rates among the Taiwan Population

Chronic obstructive pulmonary disease (COPD) causes a high disease burden among the elderly worldwide. In Taiwan, the long-term temporal trend of COPD mortality is declining, but the geographical disparity of the disease is not yet known. Nationwide COPD age-adjusted mortality at the township level during 1999–2007 is used for elucidating the geographical distribution of the disease. With an ordinary least squares (OLS) model and geographically weighted regression (GWR), the ecologic risk factors such as smoking rate, area deprivation index, tuberculosis exposure, percentage of aborigines, density of health care facilities, air pollution and altitude are all considered in both models to evaluate their effects on mortality. Global and local Moran’s I are used for examining their spatial autocorrelation and identifying clusters. During the study period, the COPD age-adjusted mortality rates in males declined from 26.83 to 19.67 per 100,000 population, and those in females declined from 8.98 to 5.70 per 100,000 population. Overall, males’ COPD mortality rate was around three times higher than females’. In the results of GWR, the median coefficients of smoking rate, the percentage of aborigines, PM₁₀ and the altitude are positively correlated with COPD mortality in males and females. The median value of density of health care facilities is negatively correlated with COPD mortality. The overall adjusted R-squares are about 20% higher in the GWR model than in the OLS model. The local Moran’s I of the GWR’s residuals reflected the consistent high-high cluster in southern Taiwan. The findings indicate that geographical disparities in COPD mortality exist. Future epidemiological investigation is required to understand the specific risk factors within the clustering areas.     

基于地理加权回归的茶叶种植专业化空间格局及影响因素——以福建省安溪县为例

以专业化茶叶种植大县安溪县为例,通过评估各乡镇茶叶种植专业化水平,结合Pearson相关分析、普通最小二乘法模型(OLS)和地理加权回归模型(GWR),筛选出平均高程、农民人均纯收入、农业从业人口比重、距交通道路距离4个主要影响安溪县茶叶种植专业化程度的因素,并探讨其对茶叶种植专业化影响程度的空间分异规律.结果表明: 安溪县各乡镇茶叶种植专业化程度呈现显著的空间自相关,以县城为中心,茶叶种植专业化程度由近及远呈现出“低-中-高”的形似杜能农业区位模型的圈层结构;GWR的拟合度(0.624)高于OLS(0.595),且前者对空间数据的解释力更高;与杜能农业区位模型的市场距离决定机制相悖,茶叶种植专业化程度受山地自然环境因素的影响明显较社会经济因素更大;茶叶种植专业化程度与平均高程、农民人均纯收入、农业从业人口比重呈正相关,而与距交通道路距离总体呈负相关;茶叶种植专业化程度与平均高程、农民人均纯收入的回归系数主要呈现出“南高北低”的空间分布特征,农业从业人口比重则呈相反的规律,而距交通道路距离则主要呈现“西南低、东北高”的空间特征.     

宁夏中宁枸杞林地粉壤土阳离子交换量空间分布预测方法对比

土壤阳离子交换量(CEC)是土壤施肥、改良的主要依据和土壤质量的评价指标,研究土壤CEC的空间分布及模型预测可为土壤养分监测、管理及精准农业实施提供科学依据。本研究以中宁枸杞林地粉壤土为对象,在自相关、交互相关等分析基础上,采用协同克里格(CoKriging)、普通最小二乘法(OLS)、地理加权回归(GWR)和随机森林(RF)模型对土壤CEC进行回归分析,比较了制图效果及模型预测精度。结果表明:中宁枸杞林地粉壤土CEC平均值为13.12 cmol·kg^-1,属中等肥力;土壤CEC的空间分布具有自相关性,并与土壤pH、有机质、黏粒和电导率在不同滞后距离上存在不同的空间相互关系; RF模型预测图避免了CoKriging、OLS和GWR模型预测图中土壤CEC图斑边界两侧破碎程度大、突变明显的缺陷,使土壤CEC在空间变化上表现为自然、平缓的过渡; RF模型RMSE值分别比CoKriging、OLS和GWR模型减少33.82%、20.55%和19.81%,R²分别提高8.84%、51.92%和7.69%。RF模型考虑了样点空间位置,明显提高了插值精度且...     

基于GWR模型的东江水质空间分异与水生态功能分区验证

流域水生态功能分区研究是我国正在开展的一项重要工作,如何验证分区结果的合理性,是当前亟待解决的问题。采用地理加权回归(GWR)模型评估流域特征对东江水质的影响,验证水质及流域影响空间差异是否与一二级水生态功能分区结果吻合,并对比了GWR模型与普通最小二乘(OLS)模型性能,讨论了GWR在分区验证方面的应用价值及不足。结果显示:1)水质指标以及GWR模型局部解释率(Local R~2)均在一二级水生态功能分区间存在显著差异;2)相比OLS模型,GWR模型校正R~2更高,残差空间自相关指数Moran′s I更低。研究表明东江水生态功能分区结果能合理反映水陆耦合关系,有效解释水质空间差异。此外建议选择总氮(TN)而非溶解氧(DO)和总磷(TP)作为分区验证指标。GWR模型在分区结果验证中具有广泛应用前景。降低数据空间自相关影响及改善距离测度方法是未来GWR模型研究的难点问题。     

基于地理加权回归拓展模型的天然次生林碳储量空间分布

为精准获取区域尺度天然次生林的碳储量及其空间分布格局,以吉林省汪清林业局浪溪林场的天然次生林为研究对象,基于165块局级固定样地,以林分因子、地形因子和土壤因子为影响因子,将普通地理加权回归模型(GWR)作为基础,从空间维度、参数异质性特征和残差空间自相关性3个方面进行改进,构建7类拓展模型,即地理海拔加权回归模型(G...     

Application of geographically weighted regression in estimating the effect of climate and site conditions on vegetation distribution in Haihe Catchment,China

Climate and topography are the two key factors influencing vegetation pattern, distribution, and plant growth. Traditionally, studies on the relationship between vegetation and climate rely largely on field data from limited samples. Now, digital elevation model (DEM) and remote sensing data readily provide huge amounts of spatial data on site-specific conditions like elevation, aspect, and climate, while recent development of geographically weighted regression (GWR) analysis facilitates efficient spatial evaluation of interactions among vegetation and site conditions. Using Haihe Catchment as a case study, GWR is applied in establishing spatial relations among leaf area index (LAI; a critical vegetation index from Moderate Resolution Imaging Spectroradiometer (MODIS)) and interpolated climate variables and site conditions including elevation, aspect, and Topographic Wetness Index (TWI). This study suggests that the GWR solution to spatial effect of climate and site conditions on vegetation is much better than ordinary least squares (OLS). In most of the study area, effects of elevation, aspect change from south to north, and precipitation on LAI are positive, while temperature, TWI, and potential evapotranspiration have a negative influence. Spatially, models perform better in places with large spatial variations in LAI—primarily driven by strong spatial variations in temperature and precipitation. On the contrary, the effect of topographic and climatic factors on vegetation is weak in regions with small spatial variations in LAI. This study shows that overall water availability is a determining factor for spatial variations in vegetation.     

Spatial patterns of species richness in New World coral snakes and the metabolic theory of ecology

The metabolic theory of ecology (MTE) has attracted great interest because it proposes an explanation for species diversity gradients based on temperature-metabolism relationships of organisms. Here we analyse the spatial richness pattern of 73 coral snake species from the New World in the context of MTE. We first analysed the association between ln-transformed richness and environmental variables, including the inverse transformation of annual temperature (1/kT). We used eigenvector-based spatial filtering to remove the residual spatial autocorrelation in the data and geographically weighted regression to account for non-stationarity in data. In a model I regression (OLS), the observed slope between ln-richness and 1/kT was ?0.626 (r² = 0.413), but a model II regression generated a much steeper slope (?0.975). When we added additional environmental correlates and the spatial filters in the OLS model, the R² increased to 0.863 and the partial regression coefficient of 1/kT was ?0.676. The GWR detected highly significant non-stationarity, in data, and the median of local slopes of ln-richness against 1/kT was ?0.38. Our results expose several problems regarding the assumptions needed to test MTE: although the slope of OLS fell within that predicted by the theory and the dataset complied with the assumption of temperature-independence of average body size, the fact that coral snakes consist of a restricted taxonomic group and the non-stationarity of slopes across geographical space makes MTE invalid to explain richness in this case. Also, it is clear that other ecological and historical factors are important drivers of species richness patterns and must be taken into account both in theoretical modeling and data analysis.     

基于GWR模型的土地覆盖与地表温度的关系——以京津唐城市群为例

基于Landsat TM土地覆盖分类数据和MODIS地表温度数据,探讨京津唐城市群不同土地覆盖的地表温度(7日),并采用常用的普通线性回归(OLS)和地理加权回归(GWR)方法分别拟合土地覆盖比例与地表温度的关系.结果表明: 研究区不同土地覆盖类型的地表温度差异明显,人工表面(40.92±3.49 ℃)和耕地(39.74±3.74 ℃)的平均温度较高,林地(34.43±4.16 ℃)和湿地(35.42±4.33 ℃)的平均温度较低;土地覆盖比例与地表温度显著相关,且两者之间的定量关系存在空间非稳定性,地理位置以及周围环境影响的差异是空间非稳定性产生的主要原因;GWR模型的拟合结果优于OLS模型(R_GWR²>R_OLS²),并且GWR模型可以量化土地覆盖比例与地表温度两者关系的空间非稳定性特征.     

福州市土壤铬含量高光谱预测的GWR模型研究

通过系统分析不同光谱分辨率和光谱变换对土壤铬高光谱预测模型的不确定性影响,筛选出最优的光谱分辨率及光谱变量进行土壤铬含量预测的地理权重回归(GWR)模型构建,利用该模型进行福州市土壤铬含量预测,并将预测结果与普通最小二乘法回归(OLS)结果进行比较分析,探讨GWR模型在土壤铬高光谱预测中的适用性及局限性。结果表明:(1)在10 nm分辨率尺度下,以土壤全铬含量为因变量,反射率的二阶微分和反射率倒数的二阶微分为自变量构建的GWR模型对土壤铬预测的效果最好。GWR模型的R~2和调节R~2分别为0.821和0.716,较OLS模型分别提高了0.529和0.450,而AIC值为720.703,较OLS模型减少了22个单位,残差平方和仅为OLS模型的1/4,说明GWR模型的预测效果较OLS模型有了显著提高。(2)土壤铬预测模型的精度受光谱分辨率影响。对于OLS预测模型来说,3 nm分辨率的模型预测效果最好,而对于GWR预测模型来说,10nm分辨率的模型不仅预测效果最好,其相较于OLS模型的改善作用显著,为土壤铬含量GWR预测的最佳光谱分辨率。(3)光谱的一阶微分变换可以有效增强土壤铬的光谱特征,而其余的光谱变换对土壤铬的光谱特征则未起到增强作用,但可以很好地提高模型的预测效果。(4)研究得出土壤铬GWR模型预测的最佳光谱分辨率为10 nm,为EO-1 Hyperion影像的光谱分辨率,而且随着采样点的增加,GWR模型的预测效果趋于稳定,适合空间异质性大的区域尺度土壤铬预测。故该模型与高光谱影像结合,实现模型从实验室尺度向区域尺度的推广,为格网尺度土壤铬的空间预测提供可能。     

Non-stationarity and local approaches to modelling the distributions of wildlife

Despite a growing interest in species distribution modelling, relatively little attention has been paid to spatial autocorrelation and non-stationarity. Both spatial autocorrelation (the tendency for adjacent locations to be more similar than distant ones) and non-stationarity (the variation in modelled relationships over space) are likely to be common properties of ecological systems. This paper focuses on non-stationarity and uses two local techniques, geographically weighted regression (GWR) and varying coefficient modelling (VCM), to assess its impact on model predictions. We extend two published studies, one on the presence–absence of calandra larks in Spain and the other on bird species richness in Britain, to compare GWR and VCM with the more usual global generalized linear modelling (GLM) and generalized additive modelling (GAM). For the calandra lark data, GWR and VCM produced better-fitting models than GLM or GAM. VCM in particular gave significantly reduced spatial autocorrelation in the model residuals. GWR showed that individual predictors became stationary at different spatial scales, indicating that distributions are influenced by ecological processes operating over multiple scales. VCM was able to predict occurrence accurately on independent data from the same geographical area as the training data but not beyond, whereas the GAM produced good results on all areas. Individual predictions from the local methods often differed substantially from the global models. For the species richness data, VCM and GWR produced far better predictions than ordinary regression. Our analyses suggest that modellers interpolating data to produce maps for practical actions (e.g. conservation) should consider local methods, whereas they should not be used for extrapolation to new areas. We argue that local methods are complementary to global methods, revealing details of habitat associations and data properties which global methods average out and miss.     

2000-2019年中国南方竹林区水分利用效率时空特征及驱动机制

水分利用效率(Water Use Efficiency, WUE)是深入理解生态系统碳、水循环及两者耦合关系的重要指标,然而我国重要森林类型之一的竹林的WUE时空格局及其驱动机制研究不足。通过MODIS净初级生产力(NPP)和蒸散(ET)数据得到竹林区WUE,采用线性趋势法计算WUE年际变化率表征变化趋势,并应用地理加权回归(GWR)模型分析了WUE与气候和地形等10个驱动因子的关系,探究了中国南方竹林区近20年间(2000—2019)WUE驱动机制。结果表明：(1)2000—2019年中国南方竹林区WUE多年均值为0.89 gC m^-2 mm^-1,呈显著下降趋势,下降速率为0.0028 gC m^-2 mm^-1 a^-1,ET上升速度大于NPP上升速度是造成WUE下降的主要原因;WUE呈南高北低的空间分布格局,83.5%区域的WUE呈下降趋势。(2)基于GWR模型的WUE驱动力分析发现,WUE变化最强的驱动因子是CO₂浓度和年降水量,而海拔、坡度等地形因子的...     

基于GWR模型的于田绿洲土壤表层盐分空间分异及其影响因子

以于田绿洲为研究靶区,利用24个采样点的土壤表层盐分数据,选取9个与土壤表层盐分密切相关的影响因子,结合空间自相关、传统回归分析和地理加权回归模型,分析表土盐分的空间分布特征及其影响因子的空间分异.结果表明:于田绿洲表土盐分在空间上并非随机分布,而是存在较强的空间依赖关系,空间自相关指数为0.479.地下水矿化度、地下水埋深、高程和温度是影响干旱区平原绿洲表土积盐的主要因子,这些因子具有空间异质性,选取的9个环境变量中除土壤pH值外,其他变量对表土盐分的影响强度均存在显著的空间分异.GWR模型对存在空间非平稳性数据的解释能力和估计精度都优于OLS模型,而且在模型估计参数的可视化上具有明显优势.