在横山县朱家沟小流域采集111个样点,分析土壤属性与地形因子相互关系,进行空间预测分析。结果表明:土壤密度与复合地形指数CTI显著正相关,土壤有机质与复合地形指数CTI、汇流动力指数SPI、沉积物运移指数STI显著负相关,全磷只与坡度β显著负相关。多元线性逐步回归模型对于土壤密度和有机质拟合较好,而对于全磷,预测结果较差;回归-克里格预测有效地减小了残差,消除了平滑效应,预测极差更宽,对于局部的变异及地形、土地利用的关系体现更为细化,提高了预测精度。
Analysis and forecast on the spatial distribution and dynamics of soil properties is an important element of sustainable land management. Topography and land use are key factors which affect soil properties variability on catchment scale in the loess hilly area. Correlation analyses were carried out between soil properties (include bulk density(BD), soil organic matter (SOM) and total phosphorus (TP)) and terrain attributes. Finally, terrain attributes and land use types were used to predict spatial distribution of soil properties by multiple-linear regression analysis and regression-Kriging, based on data from 111 points in Zhujiagou catchment; Hengshan County of Shaanxi Province. Results indicated that there is a positive correlation between bulk density and compound topographic index (CTI); and a negative correlation between SOM and compound topographic index (CTI), stream power index(SPI) and sediment transport index (STI); and a significant negative correlation was found between TP and slope(β). The correlations between these terrain attributes and soil properties reflected to some degree that soil develops in response to the way water flows through and over the landscape. The regression model for BD had the highest R2 value, followed by SOM and TP. The regression model was precise for the BD, but the variation is rather large and there was a high smoothing effect on the predicted values. For TP, the predicted result is very poor. Such techniques may be applied as a frist step in unmapped areas to guide soil sampling and model development. In order to explain the variations better, combined step-wise regression and residuals interpolated using Kriging was conducted. The results showed that regression-Kriging improved accuracy of prediction.
[1] Florinsky I V, Eilers R G, Manning G R, et al. Prediction of soil properties by digital terrain modeling[J]. Environmental Modelling & Software, 2002,17:295-311.
[2] McKenzie N J, Austin M P. A quantitative Australian approach to medium and small scale surveys based on soil stratigraphy and environmental correlation[J].Geoderma,1993,57: 329-355.
[3] Odeh I O A, McBratney A B, Chittleborough D J. Spatial prediction of soil properties from landform attributes derived from a digital elevation model[J]. Geoderma,1994,63:197-214.
[4] 张 婷,汤国安,王 春,等.黄土丘陵沟壑区地形定量因子的关联性分析[J].地理科学,2005,25(4):467~472.
[5] 张光辉,刘国彬. 黄土丘陵区小流域土壤表面特性变化规律研究[J].地理科学,2001, 21(2):118~122.
[6] 孙维侠,史学正,于东升,等.基于1:100万土壤空间数据库的有机碳储量估算研究[J].地理科学,2004,24(5):567~572.
[7] 邱 扬,傅伯杰,王 军,等.黄土丘陵小流域土壤物理性质的空间变异[J].地理学报,2002,57(5):587~594.
[8] 傅伯杰,陈利顶,马克明.黄土丘陵区小流域土地利用变化对生态环境的影响[J].地理学报,1999,54 (3):241~246.
[9] Moore I D, Gessler P E, Nieslen G A, et al. Soil attribute prediction using terrain analysis[J]. Soil Sci. Soc. Am. J.,1993,57:443-452.
[10] 连 纲,郭旭东,傅伯杰,等.黄土丘陵沟壑区县域土壤有机质空间分布特征及预测[J].地理科学进展,2006,25(2): 112~122.
[11] Tomislav H, Gerard B M, Heuvelink, et al. A generic framework for spatial prediction of soil variables based on regression-kriging[J]. Geoderma, 2004,120:75-93.
[12] Edward Batschelet. Circular statistics in biology[M]. New York:Academic Press,1981.
[13] Nyberg L. Spatial variability of soil water content in the covered catchment of Gardsjon[J]. Sweden Hydrology Processes,1996,10:89-103.
[14] 郭旭东,傅伯杰,陈利顶,等.河北省遵化平原土壤养分的时空变异特征[J].地理学报,2000,55(5):555~566.
[15] 刘付程,史学正,潘贤章,等.太湖流域典型地区土壤磷素含量的空间变异特征[J].地理科学,2003,23(1):77~81.
[16] 钟晓兰,周生路,赵其国.长江三角洲地区土壤重金属污染特征及潜在生态风险评价——以江苏太仓市为例[J].地理科学,2007,27(4):395~400.
[17] 李 翔,潘瑜春,赵春江,等.利用不同方法估测土壤有机质及其对采样数的敏感性分析[J].地理科学,2007,27(5): 689~694.
[18] 叶 玮,杨立辉,朱丽东,等.中亚热带网纹红土的稀土元素特征与成因分析[J].地理科学,2008,28(1):40~44.
[19] 赵彦锋,郭恒亮,孙志英,等.基于土壤学知识的主成分分析判断土壤重金属来源[J].地理科学,2008,28(1):45~50.
