Estimating effects of terrain attributes on local soil organic carbon content in a semi-arid pastureland
Estimating effects of terrain attributes on local soil organic carbon content in a semi-arid pastureland
复制标题
半干旱牧场地形属性对当地土壤有机碳含量的影响估算
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
M. Masihabadi
中科院分区：
文献类型：
--
作者：
S. Moghimi;M. Mahdian;Y. Parvizi;M. Masihabadi
土壤有机碳（SOC）是大气碳的来源或汇，其在土壤物理、化学和生物特性中的重要性日益得到认识。本研究的目的是预测和评价地形属性对伊朗Mereg流域丘陵牧场土壤有机碳含量的影响。地形属性包括高程、坡度、平面、剖面曲率等一级因子和坡向组合指数、湿度指数、水流功率等二级因子。采用多元线性回归（MLR）和径向基函数（RBF）人工神经网络。模型评价标准的比较表明，RBF模型（R=0.954， RMSE=0.087%）比MLR模型（R=0.528， RMSE=0.349%）更准确地预测了SOC。结果表明，隐含层有15个神经元和2个扩散值的RBF模型的预测结果比MLR模型更可靠。结果表明，土壤有机碳含量对剖面曲率、平面曲率、变换坡向和坡度百分比最为敏感。
Soil organic carbon (SOC) is a source or sink of atmospheric carbon and importance of it has been increasingly recognized in soil physical, chemical and biological characteristics. The objective of this study was to predict and evaluate the effects of topographic attributes on the soil organic carbon content at a hilly pastureland in Mereg watershed, Iran. In this research, topographic attributes include the primary factors such as elevation, slope, plan and profile curvature, transformed aspect and secondary factors such as slope-aspect combinative index, wetness index and stream power. Multiple linear regression (MLR) and radial basis function (RBF) artificial neural network were employed. The comparison of model evaluation criteria demonstrates that the RBF model (R=0.954, RMSE=0.087%) provides more accurate predictions of SOC than the MLR model(R=0.528, RMSE=0.349%). The RBF model, with 15 neurons in hidden layer and 2 spread value was applied successfully and exhibited the more reliable predictions than the MLR model. Results showed that, SOC content were mostly sensitive to the profile curvature, plan curvature, transformed aspect and slope percent.