The Moon has been divided into three terranes:Procellarum KREEP Terrane(PKT),Feldspathic Highland Terrane(FHT),and South Pole-Aitken Terrane(SPAT),using globally measured Th and FeO.Many lunar evolu-tion models have p...The Moon has been divided into three terranes:Procellarum KREEP Terrane(PKT),Feldspathic Highland Terrane(FHT),and South Pole-Aitken Terrane(SPAT),using globally measured Th and FeO.Many lunar evolu-tion models have predicted that a lunar magma ocean will produce a residual layer enriched in incompatible elements such as K,REE,and P(i.e.,KREEP)in the late age of crys-tallization;and that the distribution of thorium can be used as a proxy for determining the global distribution of KREEP.The thorium distribution in these three terranes is inhomo-geneous.The highest concentration of thorium is in PKT,the medium concentration of thorium is in SPAT,and almost none in FHT.Then what is the specific distribution in each of the terrane and what enlightenment can it tell us?Here we present and describe the detailed thorium distribution in PKT,SPAT,and FHT and provide some information for the origin of asymmetries on the lunar surface.展开更多
月球南极-艾特肯盆地是太阳系最大的撞击盆地之一,也是月球上最大、最古老的撞击盆地.南极-艾特肯盆地是研究早期大型撞击事件的重要窗口,而小型撞击坑的识别与计数定年是研究南极-艾特肯盆地演化史的基础.由于撞击坑直径和数量符合幂...月球南极-艾特肯盆地是太阳系最大的撞击盆地之一,也是月球上最大、最古老的撞击盆地.南极-艾特肯盆地是研究早期大型撞击事件的重要窗口,而小型撞击坑的识别与计数定年是研究南极-艾特肯盆地演化史的基础.由于撞击坑直径和数量符合幂次定律,数量众多的小型撞击坑难以单纯依靠人力进行识别.近年来,计算机算力的提升使得训练复杂的卷积神经网络成为可能.采用已有的专家标注训练神经网络,进而实现图像特征的自动提取,能够在保证准确率的同时极大地提高识别效率.采用基于卷积神经网络算法的You Only Look Once Version5(YOLO V5)目标探测系统来自动识别月球南极-艾特肯盆地直径为2~15 km的小型撞击坑.在训练神经网络时,使用融合了SELENE和LRO数据的数字高程模型SLDEM2015和最新的专家标记撞击坑数据库.训练好的网络在测试集上的结果与专家标记的撞击坑数据库相比,识别结果的准确率(Precision)为0.96,召回率(Recall)为0.95,F1值为0.95.通过对与专家标注不符的识别结果进行可视化,识别出至少十个专家误标记的撞击坑,证明撞击坑自动识别方法可以用于检验专家标注的可靠性.基于南极-艾特肯盆地的撞击坑自动识别结果,确定了南极-艾特肯盆地四个典型中型撞击坑的绝对模式年龄,并与已有的定年结果对比,进一步验证了自动识别结果的可靠性,也显示了提出的方法在利用自动识别的撞击坑进行中型撞击坑定年方面的潜力.提出的撞击坑自动识别方法有望进一步拓展到更小撞击坑的识别,并迁移到月球其他地质单元乃至其他行星的研究中.展开更多
基金This work was supported by National Key Research and Development Program of China(Grant No.2022YFF0503100)the Strategic Priority Research Program of Chinese Academy of Sciences(Grant No.XDB 41000000).
文摘The Moon has been divided into three terranes:Procellarum KREEP Terrane(PKT),Feldspathic Highland Terrane(FHT),and South Pole-Aitken Terrane(SPAT),using globally measured Th and FeO.Many lunar evolu-tion models have predicted that a lunar magma ocean will produce a residual layer enriched in incompatible elements such as K,REE,and P(i.e.,KREEP)in the late age of crys-tallization;and that the distribution of thorium can be used as a proxy for determining the global distribution of KREEP.The thorium distribution in these three terranes is inhomo-geneous.The highest concentration of thorium is in PKT,the medium concentration of thorium is in SPAT,and almost none in FHT.Then what is the specific distribution in each of the terrane and what enlightenment can it tell us?Here we present and describe the detailed thorium distribution in PKT,SPAT,and FHT and provide some information for the origin of asymmetries on the lunar surface.
基金中国科学院B类先导科技专项培育项目(XDB18000000)国家自然科学基金面上项目(41373068,41773065)+2 种基金国家自然科学基金重大项目(41490634)科技部科技基础性工作专项(2015FY210500)Brown University,Office of Vice President for Research SEED grant"Engaging the Chinese Lunar Exploration Program"(CLEP)
文摘月球南极-艾特肯盆地是太阳系最大的撞击盆地之一,也是月球上最大、最古老的撞击盆地.南极-艾特肯盆地是研究早期大型撞击事件的重要窗口,而小型撞击坑的识别与计数定年是研究南极-艾特肯盆地演化史的基础.由于撞击坑直径和数量符合幂次定律,数量众多的小型撞击坑难以单纯依靠人力进行识别.近年来,计算机算力的提升使得训练复杂的卷积神经网络成为可能.采用已有的专家标注训练神经网络,进而实现图像特征的自动提取,能够在保证准确率的同时极大地提高识别效率.采用基于卷积神经网络算法的You Only Look Once Version5(YOLO V5)目标探测系统来自动识别月球南极-艾特肯盆地直径为2~15 km的小型撞击坑.在训练神经网络时,使用融合了SELENE和LRO数据的数字高程模型SLDEM2015和最新的专家标记撞击坑数据库.训练好的网络在测试集上的结果与专家标记的撞击坑数据库相比,识别结果的准确率(Precision)为0.96,召回率(Recall)为0.95,F1值为0.95.通过对与专家标注不符的识别结果进行可视化,识别出至少十个专家误标记的撞击坑,证明撞击坑自动识别方法可以用于检验专家标注的可靠性.基于南极-艾特肯盆地的撞击坑自动识别结果,确定了南极-艾特肯盆地四个典型中型撞击坑的绝对模式年龄,并与已有的定年结果对比,进一步验证了自动识别结果的可靠性,也显示了提出的方法在利用自动识别的撞击坑进行中型撞击坑定年方面的潜力.提出的撞击坑自动识别方法有望进一步拓展到更小撞击坑的识别,并迁移到月球其他地质单元乃至其他行星的研究中.
