我们能否像预测天气一样预测火山喷发？
内容总结：
火山预报能否像天气预报一样精准？科学家给出乐观预期
1991年，菲律宾皮纳图博火山剧烈喷发，造成800余人死亡。幸运的是，科学家提前数日发出预警，疏散了周边25万居民，避免了更大灾难。然而，那次预报更像“有根据的猜测”，远未达到天气预报的精确度——科学家无法确定喷发的具体时间与演化路径。
三十余年过去，火山学已取得长足进步：监测仪器更先进，机器学习大幅提升数据分析效率，科学家对岩浆输送系统理解更深。如今，一个关键问题浮出水面：我们距离像预报天气那样预报火山行为还有多远？
现状：预警不等于预测
目前，即便是全球监测最完善的火山，科学家通常也只能提供“高度警惕”型预警，而非精确预测。美国地质调查局的警报系统会告知公众火山活动加剧，但这并不意味着喷发必然发生。东英吉利大学地球物理学家杰西卡·约翰逊坦言：“看起来要喷发的火山动荡，最终只有50%真的喷发。”
另一方面，有些火山偏爱“偷袭”。即使监测设备齐全，浅层高压水囊被岩浆加热后引发的蒸汽爆炸常毫无征兆。这种犹如“地雷引爆炸药库”的喷发类型，至今难以预警。
突破：频繁喷发型火山已可精准预报
在意大利斯特龙博利和埃特纳等定期喷发熔岩的火山，科学家已能提前数小时自信预报。借助地震学和地面形变测量，冰岛雷克雅内斯半岛和夏威夷基拉韦厄火山的科学家甚至可以精确追踪岩浆地下迁移路径，预测其喷发地点，误差在一小时以内。但冰岛大学火山地震学家汤姆·温德指出，这种精准预报仍“相对罕见”。
挑战：破解地下岩浆房的“黑箱”
预报火山喷发的核心难题在于，岩浆深埋地下数公里，多数活火山喷发间隔数十年，且每座火山都独一无二。温度、压力、气体含量、晶体比例、围岩强度等众多因素共同决定喷发与否。卡内基科学研究所火山学家戴安娜·罗曼坦言：“我们甚至没有完全理解背后的物理学原理。”
当前研究正从“模式识别”向“因果理解”转变。英国布里斯托尔大学领衔的“意料之外”项目，在东加勒比海地区部署数百台地震仪和光纤电缆网络，结合机器学习实时追踪岩浆活动，试图提炼出控制火山喷发的流体力学方程。
未来愿景：需要一场“地质学曼哈顿计划”
美国黄石火山观测站科学家迈克·波拉德表示：“没有理由认为未来我们无法实现像天气预报那样的火山预报。”但实现这一目标需要：
- 长期密集监测：对多样性火山群持续数十年观测，积累多周期喷发数据。目前即使美国太平洋西北地区的诸多危险火山，传感器覆盖仍不充分。
- 建立统一理论：通过海量地球物理和地球化学数据，提炼控制所有火山的共同物理方程，构建通用火山模型。
- 原位观测：冰岛克拉夫拉岩浆试验站计划钻井直抵岩浆层，实现全球首个直接岩浆观测。
“我们真正缺少的是更多数据，”德国亥姆霍兹中心地球物理学家马里乌斯·伊斯科恩表示，“尚未观测到足够多不同系统的喷发过程，但这一空白将随时间被填补。”罗曼则呼吁：“是时候全力推进了。”
每天，火山学家正以科学奇迹保护数百万人免受火山威胁。想象一下，未来人们能提前数天甚至数周撤离危险区域——这令人无比振奋。
中文翻译：
我们能否像预报天气一样预报火山喷发？
引言
1991年夏天，菲律宾的皮纳图博火山走向了自我毁灭。喷发始于6月12日，三天后，一场巨大的爆炸将活动推向高潮。当炽热的火山碎屑流——由熔岩和气体构成的发光雪崩——沿着它那已化为荒芜的山坡倾泻而下时，皮纳图博的山峰已被抹平，取而代之的是一个宽2.5公里的巨大裂口。
这场喷发导致800多人丧生，主要原因是雨水浸透的火山灰压垮了屋顶。但后果本可能严重得多：约有25万人生活在火山的阴影之下，分布在多个城市和一个庞大的美国空军基地。当皮纳图博在同年4月开始震颤并喷出蒸汽时，来自美国和菲律宾的科学家们部署了一系列仪器，追踪着火山内部的不安。
“我们对那座火山了解不多，所以进行了一次非常快速的地质评估。评估结果是：‘哦，糟了，这家伙一旦喷发，只会是大规模喷发。’”美国地质调查局黄石火山观测站的现任负责人迈克·波兰说。“这成为了预报的基础。”
到6月初，火山灰和熔岩开始从皮纳图博的山侧逸出，当局下令疏散，而仅仅几天后，灾难性的重锤便落下了。换句话说，这是一次极其惊险的脱逃。
那些科学家挽救了无数生命，但他们的预报更像是一种有依据的推测，而非表面看上去那样确凿。它完全不像天气预报；他们无法以任何接近确定性的口吻说，6月12日将会发生一次爆炸性喷发，也无法预测这次喷发将如何演变。
除了极少数例外，这种不精确性对所有受到良好监测的火山都适用。但自皮纳图博火山爆发以来，火山学作为一个领域已取得了巨大飞跃。仪器设备更加先进，机器学习使解读数据变得高效得多，科学家们对驱动火山活动的岩浆管道系统也有了更深入的理解。这促使我——一个受过专业训练、如今经常撰写该领域文章的火山学家——去思考：我们距离像预报天气那样预报火山行为，究竟还有多远？
如今，我们知道一场特定强度的风暴将在几天后降落于某个特定城市。科学家们能否有朝一日说出，一周后，某座火山有80%的概率会以某种特定方式喷发——比如熔岩喷涌、带有某种爆炸力、火山碎屑流将沿其西侧山坡而下？我四处打听，既听到了怀疑之声，也感受到了令人惊讶的乐观情绪。“简短的回答是‘能’——否则我也不会干这行了，”华盛顿特区卡内基科学研究所的火山学家戴安娜·罗曼说。
尽管人类观测天空、预测天气已有数千年历史，但当代科学的天气预报却是近代的产物：奠定这些模型基础的第一批数学方程式是在20世纪初推导出来的。如今，气象学家可以针对一个混乱无序的系统——地球的大气、海洋和陆地形态——做出长达未来两周的准确预报。
天气影响的人比火山活动更多——也就是说，它影响着所有人，且无时无刻不在影响——但约有8亿人生活在活火山100公里范围内，而一些（非常罕见的）喷发也能影响整个地球。天气和火山活动都是我们想要理解的复杂系统，但它们给预报带来的挑战却截然不同。
“火山与天气预报之间的巨大区别在于，天气一直在发生，”英国布里斯托大学的火山学家珍妮·巴克利说。大气对气象学家来说始终是可观测和可测量的。“就连他们也会说，他们需要更多的观测数据。”而岩浆则不同，它位于地壳之下数公里处，而且，大多数活火山每隔几十年才喷发一次。
每座火山也都是独一无二的。将岩浆引向地表的深层地下通道结构、岩浆的化学成分、喷发的节奏以及喷发方式的类型，都因地而异。而且，喷发并非只有一个触发因素。岩浆库的温度和压力、围岩的脆弱程度、气体和晶体的含量、岩浆的深度、区域性的板块构造运动——这些因素共同决定了火山是会猛烈爆发，还是会无果而终。
“地质学是混沌的，”德国波茨坦GFZ亥姆霍兹地球科学中心的地球物理学家马里乌斯·伊斯肯说。但混沌之中埋藏着秩序。我们能找到它吗？
我把火山想象成由数百种不同乐器组成的管弦乐队。预报喷发并非关乎听清音乐。我们已经在做了：地震仪能感知到岩浆上升时岩石的破裂；地面传感器和卫星能追踪地壳的形变，指示岩浆流向何处；气体探测器能揭示岩浆何时上升到浅层、减压并释放出有毒气体。
挑战在于，要在交响乐奏响之前很久，就知道它将如何发展至高潮。如今，在那些受到最全面监测的火山上，火山学家通常所能提供的最佳方案并非预测，而是一种高度警觉。通常，预警系统——包括美国地质调查局使用的那些——会在火山表现出加剧或升级的不安状态时通知公众。但这并不意味着喷发迫在眉睫。“看似将要喷发的火山不安状态中，只有50%最终真的导致了喷发，”英国东英吉利大学的地球物理学家杰西卡·约翰逊说。
另一方面，有些火山更喜欢打我们一个措手不及，即便它们被仪器层层包围。被困在浅层地表下的高压水囊可能被邻近的岩浆体加热。如果这个水囊破裂，就会引发危险的蒸汽爆炸，进而可能释放被禁锢的岩浆。这类喷发往往在没有任何明显预警信号的情况下发生，就像在埋藏着一座炸药山旁边引爆了一颗地雷。
如果一座火山在多个喷发周期中都经过研究，就能获得更多预测性的细节。在某些山峰，比如意大利的斯特龙博利火山和埃特纳火山，它们会定期喷出熔岩喷泉，科学家们可以自信地预报一次爆发。“我们拥有能在几小时内告知我们火山将喷发的系统，”佛罗伦萨大学的地球物理学家毛里齐奥·里佩佩说。
利用地震学和地面形变测量，其他火山（包括夏威夷的基拉韦厄火山和冰岛雷克雅未克半岛上的火山）的科学家们能够以惊人的精度追踪岩浆在地下迁移，从而确切知道岩浆将在何处以熔岩形式涌出，误差在一小时左右。但如此精确的预报“相对罕见”，冰岛大学的火山地震学家汤姆·温德说。这些是频繁活动的火山，不太可能产生重大爆炸性事件，周边社区的人们通常也知道要对它们保持警惕。在大多数其他情况下，最早的预警时间——也许在喷发前一小时左右——并不总是足以让人们撤离到安全地带。
预报喷发是一项艰巨的任务，因为火山无法简化为简单的模型。它们是结构复杂的地质怪兽，拥有隐蔽而迷宫般的管道系统。二十年前，在我读地球科学本科的第一年，一位讲师告诉我，预测下一次重大喷发的时间和地点是白日做梦——言下之意是火山太过特立独行、反复无常，彼此之间共性甚少。即使在当时，这个说法也感觉不太对劲。毕竟，它们都是承受着巨大压力和热量的容器。它们的构造图可能不同。但熔岩流经所有火山，最终，总会有东西开裂、破碎并爆炸。
我采访过的每个人都同意，科学家们仍需破解火山预报难题中一个至关重要的部分。“我们甚至还没有完全理解其背后的物理学原理，”罗曼说。是什么导致岩浆库从稳定状态转变为灾难性的失效？
“它们必定拥有共同的物理学原理，”她说。如果那些基础的方程式能被发现，或许我们可以将它们应用于所有火山，并输出一些数值，这些数值能以高精度告诉我们下一次喷发何时发生，以及其形态可能是什么样子。
科学家们已经识别出其中一些控制方程，但它们只在喷发开始后才适用。利用一个多世纪的观测数据，研究人员已在很大程度上推导出了火山灾害——特别是熔岩流和火山碎屑流——的物理学原理。例如，描述各种流体如何运动的纳维-斯托克斯方程已成功应用于这两种灾害，而热方程则揭示了这些火山流体如何以及何时冷却。如今，这些方程使得专家们能够针对特定火山预测熔岩将在何处涌出、不同类型流体的波及范围有多远，以及这一切将以多快的速度发生。
这项工作挽救了生命，但它只是预报难题的一小部分。用我们天气预报的类比来说，这就像是说：“一旦雨开始下，我们就能预测哪些流域可能会发洪水，”波兰说。而要知晓风暴何时开始，则需要深入理解岩浆库的地下物理学。
目前，喷发预警是基于识别可测量的地球物理信号（例如地震活动的升级）中出现的模式，这些信号常出现在喷发之前。但如果模式不一致（这往往是常态），仅仅靠相关性不足以进行预测。“我们试图做的是研究其中的因果关系……以理解其物理学原理，”约翰逊说。“如果你理解了这些模式意味着什么，那么当这些模式发生变化时，我们就不会那么束手无策了。”
约翰逊是一个名为“Ex-X：期待意外”的新项目成员，这是由布里斯托大学牵头的一项多学科努力，旨在调查危险火山活动升级的驱动因素。研究人员正聚焦于东加勒比地区的火山，这些火山喷发相对频繁，并且能迅速从以熔岩为主的溢流式喷发转变为突发性的、灾难性的爆炸式喷发。圣文森特岛的拉苏弗里耶尔火山就是近期的例子：2020年12月，该火山开始喷出粘稠的熔岩体，持续了数月之久。随后，多次爆炸将火山碎屑流抛向山坡。
作为这项工作的一部分，数百台地震仪以及光纤电缆网络将被用于记录哪怕是微小的地震，无论是在火山平静时期还是不安时期。这一监测工作将由机器学习程序辅助，这些程序将被训练来识别这些火山地震声景中的细微变化。近年来，这些程序已被用于处理海量数据，其熟练度和效率远超科学家独自应对的能力。这项工作已经揭示了火山下方无数以前隐藏的岩浆通道，同时也使得科学家能够近乎实时地追踪岩浆在地壳中的奔涌。
Ex-X 的理念是获得前所未有的细节，以了解岩浆行为或位置的微小变化如何导致喷发。这些洞见反过来又能阐明一些潜在的物理学原理。所有这些加勒比火山，尽管可能各具特色，但可能共享一套流体动力学方程。
然而，仅靠地震学是不够的。“我们缺乏对岩浆房里究竟发生着什么的物理学理解，”波兰说。是什么导致岩浆体中气泡的不可阻止的成核过程，这种过程能像打开汽水瓶一样，将炽热、浮力大的岩浆推穿上覆地壳？是哪种熔岩、晶体和气体的组合已经准备好触发喷发？是什么驱使喷发从涌出粘稠的熔岩转变为将火山灰和岩石喷向天空？
地球化学对此项工作也至关重要。如今，科学家们会在火山喷发期间以及喷发间隔期，在火山周围采集新鲜或古老的熔岩或火山灰，以识别化学成分的细微变化。科学家们使用复杂的数值模型来模拟火山内部情况，但这仍然是有依据的推测。不过，实验室实验或许能够为这些模型提供坚实的基础。
在实验室环境中复制最极端的现象并非易事。但在2025年秋季成功的实验中，科学家们再现了行星诞生时存在的条件，包括模拟岩浆和微型氢大气层。“你不能在地球表面凭空造出一个岩浆房，”波兰说。“但比起一段时间以前，我们现在离做到那种事可是近得太多了。”
理想情况下，火山学家们还想尝试一些真正雄心勃勃的事：“一直钻探到存在岩浆的深处，真正原位观察这些过程，而不仅仅是看到它们的结果，”温德说。这是冰岛克拉夫拉岩浆试验平台的目标之一。这个具有开创性意义的设施将成为世界上第一个直接观测岩浆的天文台。
“我们没有理由不去设想，在未来的某个时刻，我们能够拥有像天气预报那样的火山预报，”波兰说。但要推导出一套统一的火山活动理论，将需要一个地质学上的“曼哈顿计划”。
首先，需要在一系列高度多样化的火山上铺满地球物理仪器，并在多个喷发周期内进行持续监测——这意味着要持续数十年。“你可能会想，好吧，火山都得到了很好的监测。但事实并非如此，”罗曼说。“只有少数几座‘凯迪拉克’级别的火山拥有永久性监测网络。”即使是美国许多最危险的火山，例如位于太平洋西北部喀斯喀特山脉（那里有著名的圣海伦斯火山，以及岌岌可危的雷尼尔山）的火山，也只有一部分覆盖着数量有限的传感器。
有了如此海量的地球物理和地球化学信息，科学家们（在机器学习的辅助下）可以确定其共性，从而推导出基础性的地球物理定律。然后，他们就可以构建出典型的火山模型：一个非常通用的模型，但可以应用到世界上任何一座火山之上。
假设你担心日本那座容易爆炸的富士山。科学家们可以将其当前的地震活动状态、岩浆地球化学特征以及变形速率输入到模型中。然后，由这些控制方程驱动的软件可以虚拟地快进火山演变过程，直达其最可能的喷发日期，同时描述最可能的喷发方式和持续时间。
一些专家推测，可能存在几种火山原型——例如，偏好喷出熔岩的类型，或者特别容易爆炸的类型。无论哪种方式，这种喷发预报概念都得到了几位火山学家的支持。“这绝对是正确的思考方向，”加州理工学院的地球物理学家和机器学习研究员扎克·罗斯说。
但对准确预报的怀疑依然存在。“目前，我只能想象在特殊情况下才能做到，”温德说，他举例提到了那些喷发频率很高的火山，比如夏威夷或冰岛的火山。
我采访过的其他人则更为乐观，他们认为，虽然某些火山总会带来麻烦——例如那些每隔几个世纪才喷发一次的火山，或者那些似乎能在几小时内从沉寂转为爆发的火山——但许多喷发应该是可以预报的。“我们真正缺少的是更多的数据。我们实际上并没有观测过那么多不同的系统爆发，”伊斯肯说。“但我认为随着时间的推移，这个缺口将会被填补。”
罗曼是正在开发中的“四维俯冲带”项目（简称SZ4D）的成员。如果能获得充足资金，这项国际性努力将在各个俯冲带——即一个构造板块俯冲到另一个板块下方的大片区域，包括智利、阿拉斯加和喀斯喀特山脉的站点——开展密集的监测活动，以研究重大山体滑坡、地震和喷发的触发因素。她希望，导致这些灾害背后的物理学原理将会显现出来。
SZ4D 将是一项庞大的科学事业。但同样巨大的努力也曾是理解天气如何运作以及地球气候如何快速变化所必需的。总得有个开始。“是时候大力推进了，”罗曼说。
每一天，火山学家都在创造科学奇迹，保护数百万计的人们免受喷发之害。对我而言，想象这样一个未来令人激动：人们不仅有几个小时，而是有几天甚至几周的时间来撤离危险地带。
英文来源：
Will We Ever Be Able To Forecast Volcanic Eruptions Like Weather?
Introduction
In the summer of 1991, Pinatubo, a volcano in the Philippines, self-destructed. The eruption started on June 12, and three days later it culminated in a tremendous explosion. By the time pyroclastic flows — incandescent avalanches of molten rock and gas — tumbled down its sterilized slopes, Pinatubo’s peak had been obliterated and replaced by a 2.5-kilometer-wide chasm.
The eruption killed more than 800 people, mainly because roofs, weighed down by rain-saturated ash, collapsed. But it could have been so much worse: About 250,000 people, across multiple cities and a sprawling U.S. Air Force base, lived in the volcano’s shadow. When Pinatubo started convulsing and belching steam in April of that year, scientists from the United States and the Philippines deployed an array of instruments that tracked the volcano’s inner tumult.
“We didn’t know much about that volcano, and so there was this really rapid geological assessment. And the assessment said, ‘Oh, crap, when this thing erupts, it only erupts big,’” said Mike Poland, current scientist in charge at the U.S. Geological Service’s Yellowstone Volcano Observatory. “And that became the basis for a forecast.”
By early June, ash and lava were escaping Pinatubo’s flanks, and an evacuation was ordered, just a few days before the cataclysmic hammer fell. It was, in other words, a very close call.
Those scientists saved countless lives, but their forecast was more of an educated guess than it might have appeared. It was nothing like a weather forecast; they couldn’t say that on June 12, an explosive eruption was going to occur with anything resembling certainty, nor could they predict the evolution of that eruption.
With very few exceptions, this imprecision is true of all well-monitored volcanoes. But volcanology, as a field, has made great leaps since Pinatubo blew its top. The instrumentation is more advanced, machine learning has made interpreting data far more efficient, and scientists have a much better understanding of the magmatic plumbing that drives volcanism. That’s prompted me — as a professionally trained volcanologist who now writes a lot about the field — to wonder: How close are we to forecasting volcano behavior the way we forecast the weather?
Today, we know that a storm of a certain magnitude will fall on a specific city in a few days’ time. Will scientists ever be able to say that a week from now, a certain volcano has an 80% chance of erupting in a particular way — with lava gushing, with a certain explosive force, with pyroclastic flows that will travel down its western flank? I asked around, and I found both skepticism and a surprising degree of optimism. “The short answer — otherwise I wouldn’t be doing this — is yes,” said Diana Roman, a volcanologist at Carnegie Science in Washington, D.C.
Though sky watchers have anticipated the weather for millennia, contemporary scientific prediction of weather is a recent invention: The first mathematical equations grounding these models were derived at the start of the 20th century. Today, meteorologists can take a pandemoniac system — Earth’s atmosphere, oceans, and landforms — and make accurate forecasts up to two weeks into the future.
Weather affects more people than volcanism — namely, everyone, all the time — but some 800 million people live within 100 kilometers of an active volcano, and some (very rare) eruptions can also affect the entire planet. Both weather and volcanism are complex systems that we want to understand, but the problems they present for forecasting are different.
“The big difference between [volcanoes] and the weather forecasting is the weather is always happening,” said Jenni Barclay, a volcanologist at the University of Bristol in England. The atmosphere is perpetually visible and measurable to meteorologists. “Even they would say they need more observations.” Magma, on the other hand, resides kilometers below Earth’s crust, and at most, active volcanoes erupt once every few decades.
Each volcano is also unique. The architecture of the subterranean pathways that funnel magma to the surface, the chemistry of the magma, the cadence of eruptions, and the assortment of eruption styles differ from place to place. And eruptions don’t have just one trigger. The temperature and pressure of the magma reservoir, the weakness of the enclosing rock, the gas and crystal content, the depth of the magma, the regional motion of tectonic plates — these factors all contribute to whether a paroxysm happens or fizzles out.
“Geology is chaotic,” said Marius Isken, a geophysicist at the GFZ Helmholtz Center for Geosciences in Potsdam, Germany. But there is order buried in the chaos. Can we find it?
I imagine volcanoes as orchestras composed of hundreds of different instruments. Forecasting eruptions isn’t about hearing the music. We already do that: Seismometers sense the cracking of rock as magma ascends; ground sensors and satellites can track shifts in the crust, indicating where magma is flowing; gas detectors reveal when magma rises to shallow depths, depressurizes, and emits noxious fumes.
The challenge comes in knowing how the symphony will develop to a climax, long before it gets underway. Today, at the most comprehensively monitored volcanoes, the best that volcanologists can normally offer is not prediction but a form of acute caution. Often, alert systems — including those used by the U.S. Geological Survey — notify the public if a volcano is exhibiting heightened or escalating unrest. But that doesn’t mean an eruption is imminent. “Only 50% of volcanic unrest that looks like it’s going to be an eruption ends up in an eruption,” said Jessica Johnson, a geophysicist at the University of East Anglia in England.
On the other hand, some volcanoes prefer to ambush us, even when smothered in instrumentation. Pockets of highly pressurized water trapped just below the surface can be heated by adjacent bodies of magma. If that pocket ruptures, a dangerous steam explosion follows, which can then unleash imprisoned magma. This type of eruption often occurs with no discernible warning signs, and it’s like a land mine going off next to a buried mountain of dynamite.
More predictive detail can come if a volcano has been studied over the course of several eruption cycles. At certain peaks, such as Italy’s Stromboli and Etna volcanoes, which regularly spout fountains of lava, scientists can confidently forecast an outburst. “We have systems that can tell us that in a few hours, the volcano will erupt,” said Maurizio Ripepe, a geophysicist at the University of Florence.
Using seismology and ground deformation measurements, scientists at other volcanoes, including Hawai‘i’s Kīlauea and those on Iceland’s Reykjanes Peninsula, can track magma migrating underground with such staggering precision that they know exactly where it will emerge as lava, to within an hour or so. But such precise forecasts are “relatively unusual,” said Tom Winder, a volcano seismologist at the University of Iceland. These are frequently active volcanoes, unlikely to produce a major explosive event, and people in surrounding communities generally know to be wary of them. In most other cases, the earliest warning times — perhaps an hour or so before the eruption — aren’t always enough to get people to safety.
Forecasting eruptions is a big ask because volcanoes cannot be reduced to simple models. They’re baroque geologic beasts with hidden, labyrinthine plumbing. Twenty years ago, during my first year as a geoscience undergraduate, a lecturer told me that predicting when and where the next major eruption would take place was a pipe dream — the implication being that volcanoes are far too idiosyncratic and mercurial to have much in common with one another. That comment felt off even then. After all, they are all vessels of immense pressure and heat. Their schematics may differ. But molten rock flows through all of them, and eventually, something cracks, breaks, and explodes.
Everyone I spoke to agreed that scientists still need to crack a vital piece of the volcano forecasting puzzle. “We don’t even fully understand the underlying physics,” Roman said. What causes a magma reservoir to transition from a stable state to catastrophic failure?
“They have to have shared physics,” she said. If those underlying equations can be discovered, perhaps we can apply them to all volcanoes and output values that tell us, with high accuracy, when the next eruption is due, and what its shape may be.
Scientists have already identified some of these governing equations, but they only apply after eruptions have begun. Using more than a century of observations, researchers have largely derived the physics of volcanic hazards — particularly lava flows and pyroclastic flows. For examples, the Navier-Stokes equations, which describe how fluids of all kinds move, have been successfully applied to both of these hazards, while the heat equation reveals how and when these volcanic fluids cool down. Today, they allow experts to predict, for specific volcanoes, where outpourings will emerge, how far the different kinds of flows will reach, and how quickly it will all happen.
This work saves lives, but it’s a fraction of the forecasting dilemma. Using our weather analogy, this is like saying, “Once the rain starts to fall, we can forecast what watersheds might flood,” Poland said. Knowing when the storm will start requires getting at the subsurface physics of magma reservoirs.
For now, eruption warnings are based on recognizing patterns in measurable geophysical signals, such as an escalation of seismic activity, that precede eruptions. But correlation isn’t enough for prediction if the patterns aren’t consistent, which is often the case. “What we’re trying to do is looking at the causative relationships there … to understand the physics,” Johnson said. “If you understand what those patterns mean, [then] when those patterns change, we’re not that stuck.”
Johnson is part of a new project named Ex-X: Expecting the Unexpected, a multidisciplinary effort led by the University of Bristol to investigate the drivers of dangerous volcanic escalations. Researchers are focusing on the volcanoes of the Eastern Caribbean, which erupt relatively frequently and can quickly transition from effusive-style, lava-heavy eruptions to sudden, catastrophic, explosive ones. La Soufrière, on the island of St. Vincent, provided a recent example of this: In December 2020, the volcano began expelling a viscous mass of lava, which continued for several months. Then multiple explosions threw pyroclastic flows down its slopes.
As part of this work, hundreds of seismometers, as well as networks of fiber-optic cables, will be used to record even the tiniest of earthquakes, during periods of tranquility and unrest. This monitoring effort will be aided by machine learning programs that will be taught to identify minute shifts in the seismic soundtrack of these volcanoes. In recent years, these programs have been used to process a huge volume of data far more proficiently and efficiently than scientists can manage alone. This work has already revealed myriad previously hidden magmatic pathways beneath volcanoes while also permitting scientists to track, almost in real time, magma barreling through the crust.
The idea of Ex-X is to gain unprecedented detail on how tiny changes in the behavior or position of magma can lead to eruptions. Those insights can, in turn, illuminate some of the underlying physics. All these Caribbean volcanoes, diverse though they may be, could have a shared set of fluid dynamics equations.
However, seismology won’t be enough by itself. “We lack the physical understanding of what exactly is going on in a magma chamber,” Poland said. What causes the unstoppable nucleation of bubbles within a body of magma, which can propel hot, buoyant magma through the crust above with soda can–like effervescence? What combination of molten rock, crystals, and gas is primed to trigger an eruption? What drives an eruption to switch from expelling oozing lava to blasting ash and rock into the sky?
Geochemistry is essential to this effort, too. Today, scientists scoop up lava or ash, fresh or ancient, around volcanoes — both during an eruption and in the interregnum between them — to identify subtle changes in chemical makeup. Scientists use sophisticated numerical models to simulate volcanic viscera, but this is still educated guesswork. Laboratory experiments, though, may be able to ground these models.
Replicating the most extreme phenomena in laboratory settings is not easy. But in successful experiments in the fall of 2025, scientists re-created the conditions present at the birth of planets, complete with simulacra of magma and miniature hydrogen atmospheres. “You can’t just make a magma chamber at the surface of the Earth,” Poland said. “But we’re a heck of a lot closer to that sort of thing than we were a while ago.”
Ideally, volcanologists want to try something else truly ambitious: “Drill all the way down to where there is some magma sitting at depth, and really see these processes in situ, rather than just seeing the results of them,” Winder said. That is one of the objectives of the Krafla Magma Testbed in Iceland. This literally groundbreaking facility is set to become the world’s first direct magma observatory.
“There’s no reason we can’t think that at some point in the future, we can have volcano forecasts that are like weather forecasts,” Poland said. But deriving a unified theory of volcanism will require a geologic Manhattan Project.
First, a constellation of highly diverse volcanoes will need to be slathered in geophysical instrumentation and consistently monitored over multiple eruption cycles — meaning many decades. “You would like to think, OK, volcanoes are pretty well monitored. But they’re not,” Roman said. “There’s a handful of Cadillac volcanoes that have permanent networks.” Even many of the United States’ most dangerous volcanoes, along the Cascades in the Pacific Northwest (home to the notorious Mount St. Helens, for example, and the precarious Mount Rainier), are only partly covered in a limited number of sensors.
With such a torrent of geophysical and geochemical information, scientists (aided by machine learning) can determine the commonalities that would allow them to derive foundational geophysical laws. Then they can build the archetypal volcano model: one that is very generic but can be layered onto any volcano in the world.
Let’s say you’re concerned about Japan’s explosion-prone Mount Fuji. Scientists could feed its current state of seismicity, its magmatic geochemistry, and the rate at which it’s deforming into the model. Software driven by those governing equations could then virtually fast-forward the volcano toward its most probable eruption date, while also describing the likeliest eruption style and duration.
Some experts suspect that there may be several volcano archetypes — ones that prefer to throw out lava, for instance, or the especially explosive kind. Either way, this eruption forecasting concept finds favor with several volcanologists. “That’s definitely the right way to be thinking about it,” said Zach Ross, a geophysicist and machine learning researcher at the California Institute of Technology.
But skepticism about accurate forecasting remains. “At the moment, I can only imagine it in exceptional circumstances,” Winder said, citing volcanoes that erupt with great frequency, like those in Hawai‘i or Iceland.
Other people I spoke with are more sanguine, suggesting that while certain volcanoes will always be troublesome — those that erupt once every few centuries, for example, or those that seem to go from silent to violent in a matter of hours — many eruptions should be forecastable. “What we’re really missing is more data. We have not really observed that many different systems going off,” Isken said. “But I think that gap will fill over time.”
Roman is part of the in-development Subduction Zones in Four Dimensions project, or SZ4D. If sufficiently funded, this international effort will carry out an intense monitoring campaign along various subduction zones — vast areas where one tectonic plate dives underneath another, including sites in Chile, Alaska, and the Cascades — to study the triggers of major landslides, earthquakes, and eruptions. She hopes that the underlying physics leading to each of these hazards will emerge.
SZ4D would be a colossal scientific undertaking. But similarly mammoth endeavors were needed to understand how the weather works, and how Earth’s climate is rapidly changing. You’ve got to start somewhere. “It’s time for a big push,” Roman said.
Every day, volcanologists perform scientific miracles to protect millions of people from eruptions. It’s thrilling for me to imagine a future in which people get not just hours, but days or even weeks to get themselves out of harm’s way.