This paper reports SHRIMP zircon U-Pb ages of 196±2 Ma for granite, and 195±1 Ma for gabbro from the Xialan complex in the Meizhou area, northern Guangdong Province. These results shed new light on the calm ...This paper reports SHRIMP zircon U-Pb ages of 196±2 Ma for granite, and 195±1 Ma for gabbro from the Xialan complex in the Meizhou area, northern Guangdong Province. These results shed new light on the calm stage of magmatic activity in southeastern China during 200-180 Ma, and revealed that the back-arc extension induced by the subduction of the western Pacific plate may have begun at 195 Ma at least. Field observation on the fresh outcrops allows us to recognize some features formed by magma mixing. A part of the gabbro has a fine-grained rim of 20-30 cm at the margin, and thins gradually to-ward the granite; numerous dark fine-grained to microcrystalline dioritic enclaves developed in the granite. These enclaves vary in shape and size, dark minerals concentrated at the margin of enclaves, and the contact between enclaves and host rock is either obvious or obscure, or gradational. In addi-tion, needle-shaped apatites are included in the enclaves. The REE patterns of gabbros, as well as the trace element patterns, are generally consistent with those of granitic rocks. The above characters further suggest that in the Early Jurassic the injection of basic magma had melted deep continental crust and produced acidic magma, and the Xialan complex was produced by the mixing of them.展开更多
滑坡空间易发性统计模型的构建需要正样本(滑坡点)和负样本(非滑坡点)两类数据,但历史观测数据仅记录了正样本,而负样本的选取容易受到正样本污染,因为没有滑坡记录的地方也可能在过去或未来发生滑坡,从而导致模型的预测精度与稳定性受...滑坡空间易发性统计模型的构建需要正样本(滑坡点)和负样本(非滑坡点)两类数据,但历史观测数据仅记录了正样本,而负样本的选取容易受到正样本污染,因为没有滑坡记录的地方也可能在过去或未来发生滑坡,从而导致模型的预测精度与稳定性受到影响。针对此问题,将前期提出的半监督学习算法PBLC(positive and background learning with constraints)应用于滑坡空间易发性分析,探讨其解决负样本污染问题的有效性。本文以粤东地区为研究区,选择高程、坡度、坡向、剖面曲率、距离道路最短距离、距离断层线最短距离、距水系最短距离、年平均降雨量、归一化植被指数和地理坐标共11个影响因子作为环境变量。结果表明,与传统的人工神经网络模型相比,基于PBLC算法的预测概率取值范围更为合理,预测结果更加稳定,且预测精度随背景样本数量增加而提高;粤东地区的滑坡灾害高易发区集中于北部和西南区域,坡度和高程是影响该地区滑坡易发性的主要因子。结果表明,半监督学习算法PBLC可以有效解决滑坡统计建模过程负样本污染的问题,提高模型预测精度。展开更多
基金Supported by National Natural Science Foundation of China (Grant Nos. 40642012, 40772134)
文摘This paper reports SHRIMP zircon U-Pb ages of 196±2 Ma for granite, and 195±1 Ma for gabbro from the Xialan complex in the Meizhou area, northern Guangdong Province. These results shed new light on the calm stage of magmatic activity in southeastern China during 200-180 Ma, and revealed that the back-arc extension induced by the subduction of the western Pacific plate may have begun at 195 Ma at least. Field observation on the fresh outcrops allows us to recognize some features formed by magma mixing. A part of the gabbro has a fine-grained rim of 20-30 cm at the margin, and thins gradually to-ward the granite; numerous dark fine-grained to microcrystalline dioritic enclaves developed in the granite. These enclaves vary in shape and size, dark minerals concentrated at the margin of enclaves, and the contact between enclaves and host rock is either obvious or obscure, or gradational. In addi-tion, needle-shaped apatites are included in the enclaves. The REE patterns of gabbros, as well as the trace element patterns, are generally consistent with those of granitic rocks. The above characters further suggest that in the Early Jurassic the injection of basic magma had melted deep continental crust and produced acidic magma, and the Xialan complex was produced by the mixing of them.
文摘滑坡空间易发性统计模型的构建需要正样本(滑坡点)和负样本(非滑坡点)两类数据,但历史观测数据仅记录了正样本,而负样本的选取容易受到正样本污染,因为没有滑坡记录的地方也可能在过去或未来发生滑坡,从而导致模型的预测精度与稳定性受到影响。针对此问题,将前期提出的半监督学习算法PBLC(positive and background learning with constraints)应用于滑坡空间易发性分析,探讨其解决负样本污染问题的有效性。本文以粤东地区为研究区,选择高程、坡度、坡向、剖面曲率、距离道路最短距离、距离断层线最短距离、距水系最短距离、年平均降雨量、归一化植被指数和地理坐标共11个影响因子作为环境变量。结果表明,与传统的人工神经网络模型相比,基于PBLC算法的预测概率取值范围更为合理,预测结果更加稳定,且预测精度随背景样本数量增加而提高;粤东地区的滑坡灾害高易发区集中于北部和西南区域,坡度和高程是影响该地区滑坡易发性的主要因子。结果表明,半监督学习算法PBLC可以有效解决滑坡统计建模过程负样本污染的问题,提高模型预测精度。
