准确获取土壤性质的空间分布信息,是区域土壤资源优化利用和土壤环境保护的需要.以川中丘陵区三台县为案例区,运用人工神经网络模型,构建融合区域定性及定量辅助变量的空间预测方法,模拟三台县土壤有机质的空间分布格局.结果表明,研究...准确获取土壤性质的空间分布信息,是区域土壤资源优化利用和土壤环境保护的需要.以川中丘陵区三台县为案例区,运用人工神经网络模型,构建融合区域定性及定量辅助变量的空间预测方法,模拟三台县土壤有机质的空间分布格局.结果表明,研究区土壤有机质在4.20～47.60 g kg-1之间,平均为17.97 g kg-1;变异系数为36.89％,属中等程度变异.土壤有机质的块金值与基台值之比为0.742,变程为7.0 km,即空间自相关性较弱.不同土壤类型间有机质含量差异显著;土属的空间分布较土类能更好地揭示研究区土壤有机质含量空间分布格局的差异.除土壤类型因素的影响外,坡度、地形湿度及植被盖度是研究区土壤有机质空间变异的主要因子.融合土壤类型因素和地形植被因子的神经网络模型预测结果,比普通克里格法、回归克里格法以及神经网络结合普通克里格的方法,更符合研究区地学规律和实际情况;其预测结果的平均绝对误差、平均相对误差和均方根误差较其他3种方法均降低幅度显著.同时,该方法对极值有较好的预测能力.研究为复杂环境条件下准确获取区域土壤性质的空间分布信息提供了较可行的方法.展开更多
利用GPS和GIS技术,研究了江西省兴国县潋水河流域小、中和大3种尺度下土壤氮素的空间分布特征。结果表明,土壤全氮和有效氮在小、中和大3种尺度下呈正态或对数正态分布,但二者均值随研究尺度的扩大而增加。其中,土壤全氮平均含量分别为0...利用GPS和GIS技术,研究了江西省兴国县潋水河流域小、中和大3种尺度下土壤氮素的空间分布特征。结果表明,土壤全氮和有效氮在小、中和大3种尺度下呈正态或对数正态分布,但二者均值随研究尺度的扩大而增加。其中,土壤全氮平均含量分别为0.60 g kg-1、0.73 g kg-1和0.83 g kg-1,有效氮含量为64.8mg kg-1、66.3 mg kg-1和80.2 mg kg-1。这3种尺度区土壤全氮和有效氮含量的空间变异函数参数和空间分布特征并不相同。因此,二者空间变异表现出明显的尺度效应。网格抽样和分层抽样相结合的采样方法均适合丘陵区小、中和大3种尺度下土壤氮素分布的研究。它们的合理采样数在90%置信度下分别为原采样方案样点数的82.6%、71.2%和58.1%,在95%置信度下分别是原采样方案样点数的95.7%、94.2%和89.2%。展开更多
Plant biomass and biodiversity, element accumulation and return, water loss and soil erosion, and changesin soil properties were studied for up to 10 years after conversion of sparse tree-shrubby grass land into thefo...Plant biomass and biodiversity, element accumulation and return, water loss and soil erosion, and changesin soil properties were studied for up to 10 years after conversion of sparse tree-shrubby grass land into thefollowing four land use patterns: masson pine (Pznvs massonzana Lamb.) land, beautiful sweetgum (Ltq-uidambar fomosana Hance) land, vegetation reservation land, and artificial mowing land. The annualbiomass production of the masson pine land was 5060 kg ha￣-1 being 4.9, 2.1, and 6.0 times that of the beau-tiful sweetgum land, the vegetation reservation land, and the artificial mowing land, respectively. Comparedwith the background values, the number of plant species for the vegetation reservation land increased by 10species after 10 years of land utilization, while for the masson pine and the beautiful sweetgum decreased by4, and for the artificial mowing land by 9. For masson pine land, total amount of N, P, K, Ca, and Mg neededfor producing 1000 kg dry matter was only 3.5 kg, annual element return through litter was 22 kg ha￣-1, bothof which were much lower than those of the other patterns. Vegetation reservation was an effective measureto conserve soil and water and improve soil fertility in the red soil hilly region. Artificial mowing arousedserious degradation of vegetation and soil. Some measures and suggestions for management and exploitationof the red soil hilly region such as masson pine planting, closing hills for afforestation, and stereo-agricultureon one hill are proposed.展开更多
文摘准确获取土壤性质的空间分布信息,是区域土壤资源优化利用和土壤环境保护的需要.以川中丘陵区三台县为案例区,运用人工神经网络模型,构建融合区域定性及定量辅助变量的空间预测方法,模拟三台县土壤有机质的空间分布格局.结果表明,研究区土壤有机质在4.20～47.60 g kg-1之间,平均为17.97 g kg-1;变异系数为36.89％,属中等程度变异.土壤有机质的块金值与基台值之比为0.742,变程为7.0 km,即空间自相关性较弱.不同土壤类型间有机质含量差异显著;土属的空间分布较土类能更好地揭示研究区土壤有机质含量空间分布格局的差异.除土壤类型因素的影响外,坡度、地形湿度及植被盖度是研究区土壤有机质空间变异的主要因子.融合土壤类型因素和地形植被因子的神经网络模型预测结果,比普通克里格法、回归克里格法以及神经网络结合普通克里格的方法,更符合研究区地学规律和实际情况;其预测结果的平均绝对误差、平均相对误差和均方根误差较其他3种方法均降低幅度显著.同时,该方法对极值有较好的预测能力.研究为复杂环境条件下准确获取区域土壤性质的空间分布信息提供了较可行的方法.
文摘利用GPS和GIS技术,研究了江西省兴国县潋水河流域小、中和大3种尺度下土壤氮素的空间分布特征。结果表明,土壤全氮和有效氮在小、中和大3种尺度下呈正态或对数正态分布,但二者均值随研究尺度的扩大而增加。其中,土壤全氮平均含量分别为0.60 g kg-1、0.73 g kg-1和0.83 g kg-1,有效氮含量为64.8mg kg-1、66.3 mg kg-1和80.2 mg kg-1。这3种尺度区土壤全氮和有效氮含量的空间变异函数参数和空间分布特征并不相同。因此,二者空间变异表现出明显的尺度效应。网格抽样和分层抽样相结合的采样方法均适合丘陵区小、中和大3种尺度下土壤氮素分布的研究。它们的合理采样数在90%置信度下分别为原采样方案样点数的82.6%、71.2%和58.1%,在95%置信度下分别是原采样方案样点数的95.7%、94.2%和89.2%。
文摘Plant biomass and biodiversity, element accumulation and return, water loss and soil erosion, and changesin soil properties were studied for up to 10 years after conversion of sparse tree-shrubby grass land into thefollowing four land use patterns: masson pine (Pznvs massonzana Lamb.) land, beautiful sweetgum (Ltq-uidambar fomosana Hance) land, vegetation reservation land, and artificial mowing land. The annualbiomass production of the masson pine land was 5060 kg ha￣-1 being 4.9, 2.1, and 6.0 times that of the beau-tiful sweetgum land, the vegetation reservation land, and the artificial mowing land, respectively. Comparedwith the background values, the number of plant species for the vegetation reservation land increased by 10species after 10 years of land utilization, while for the masson pine and the beautiful sweetgum decreased by4, and for the artificial mowing land by 9. For masson pine land, total amount of N, P, K, Ca, and Mg neededfor producing 1000 kg dry matter was only 3.5 kg, annual element return through litter was 22 kg ha￣-1, bothof which were much lower than those of the other patterns. Vegetation reservation was an effective measureto conserve soil and water and improve soil fertility in the red soil hilly region. Artificial mowing arousedserious degradation of vegetation and soil. Some measures and suggestions for management and exploitationof the red soil hilly region such as masson pine planting, closing hills for afforestation, and stereo-agricultureon one hill are proposed.
