摘要
结合地理信息系统(GIS),在面积约为2000km2的新疆典型盐渍化区渭干河流域布设了土壤取样点43个,测定了土壤表层(0~30cm)养分(全氮、全磷、全钾、碱解氮、速效磷、速效钾)的空间变异规律。结果表明:七种养分元素中除了碱解氮、速效磷的含量服从对数正态分布,其它养分元素均符合正态分布;半方差分析得出各项目都能很好的用模型来拟合,全氮、全磷、速效钾符合指数模型,碱解氮、速效磷、全磷符合球状模型。只有有机质是符合高斯模型;全氮和碱解氮的系统空间相关性很弱,而其余养分元素均表现出中等强度的空间相关性,变程在14.3km^67.0km之间,相差比较大;用Kriging插值法对未测点的养分元素进行最优估计,绘制含量分布图,从而可以更直观的反映研究区土壤养分的空间结构和分布特征。
Geostatistics combined with GIS was applied to collect forty - three samples and analyze the spatial variability of soil nutrients (total nitrogen, total phosphorus, total kalium, available nitrogen, available phosphorus and available kalium) in topsoil (0 -30cm) from about 2000km^2 of typical saline soil of Weigan drainage area in Xinjiang. The results indicated that except for available nitrogen and available phosphorus showing lognormal distribution, the others show normal distribution. The chemical analysis of semivariograms indicated that total nitro- gen, total phosphorus and available kalium are best described by exponential model, total kalium, available nitrogen and available phosphorus are best described by spherical model, only organic matter is fitted well using the Gauss model. The results also indicated that the spatial correlation of tested parameters, except total nitrogen and available nitrogen are weekly dependent, the others are moderately dependent, but their spatial heterogeneity degree was different. The ranges of soil nutrients were 14.3km - 67.0km. The spatial variability of all soil nutri- ents are mainly produced by structural factors and random factors. The Kriging method was applied to estimate the unobserved points and then the precise contour maps were obtained. By this way, the spatial variability of soil nutrients can be reflected directly.
出处
《干旱区资源与环境》
CSSCI
CSCD
北大核心
2007年第11期113-117,共5页
Journal of Arid Land Resources and Environment
基金
国家自然科学基金项目(20267002)资助