过去半个世纪，月球被视为静态、无空气且几乎无水的岩石，但轨道仪器与机器人任务显示其地质状态多样且仍缺乏约束。NASA 的 Artemis 计划规划分阶段重返：Artemis II 与 Artemis III 在轨道飞行，而 Artemis IV 将在阿波罗时代以来首次载人登月。该计划的明确目标是持续驻留，借此建立持续的资料流与广泛样本回收。未来 10 到 20 年，Artemis 可逐步取代 Apollo 50 年前的样本框架，改以反复测量和现代分析来处理遗留的月球问题。

主流巨撞假说仍主张约 45 亿年前，一颗火星大小的天体撞击原始地球，但这一推论建立在模拟与有限的 Apollo 材料之上。新任务旨在回收撞击坑与冲击区中新鲜未改造的深部岩石，以改进月球熔融洋年代表与内部演化重建。水的研究已从「没有」转为「存在」：南极永久阴影坑中有冰，表面矿物中亦有结合态水，但总量与可取用性仍未明朗。若冰以集中且较纯净的沉积存在，就可能转化为氧气与燃料；若与月壤混杂过度，早期基地的开采将可能成本过高而不可行。

内部结构、远侧与近侧的不对称及磁场仍是主要盲点。Apollo 地震仪只在有限区域采样，因此核心大小、地函结构与热分布仍是估算值。Artemis 的表面作业可在新地形部署密集地震与地球物理网络，显著提高解析度。透过比较崎岖的远侧与较平顺的近侧样本，可能揭示其年代、组成与热演化差异，并解释为何双半球长期分化。来自多地点、可精准定年的岩石与磁场测量，亦应该界定月球地磁发电机何时运作及其强度，将月球科学从未解之谜清单推进到一致的行星演化模型。

For more than half a century, the Moon was treated as a static, airless, nearly waterless rock, yet orbital instruments and robotic missions have shown it is geologically diverse and still poorly constrained. NASA’s Artemis program plans a staged return: Artemis II and III are orbital, while Artemis IV will place astronauts on the surface for the first time since the Apollo era. The program’s explicit goal is sustained presence, which could create continuous data flow and broad sample return. Over the next 10 to 20 years, Artemis could replace Apollo’s 50-year-old sample context with repeated measurements and modern analysis to tackle legacy lunar questions.

The leading giant-impact theory still assumes a Mars-sized impactor struck proto-Earth about 4.5 billion years ago, but it rests on simulations and limited Apollo material. New missions aim to recover deep, unaltered rocks from craters and impact sites to refine lunar magma-ocean chronology and internal evolution. Water research has moved from “none” to “present”: ice exists in permanently shadowed south-polar craters and bound water is trapped in surface minerals, yet abundance and accessibility remain unknown. If ice occurs in concentrated, purer deposits, it could be processed for oxygen and fuel; if it is too mixed with regolith, extraction might be uneconomic for early bases.

Internal structure, far-side asymmetry, and magnetism are still major blind spots. Apollo seismometers sampled a limited region, so core size, mantle architecture, and heat distribution are still approximations. Artemis surface operations could deploy dense seismic and geophysical networks across new terrains to improve resolution sharply. Sampling the rough far side against the smoother near side may reveal age, composition, and thermal evolution differences that explain why the hemispheres diverged for so long. Better dated rocks and magnetic measurements from multiple locations should also constrain when a lunar dynamo operated and how strong it was, helping move lunar science from a list of mysteries to a coherent model of planetary evolution。