地震科學：破壞性餘震

原文網址：www.sciencedaily.com/releases/2018/08/180829143753.htm

地震科學：破壞性餘震

研究利用人工智慧科技而開始試著對餘震發生的地點進行預測

在大地震發生過後的數周或數月之間，鄰近地區往往會受到強力餘震襲擊，撼動已經受損的社會群落並嚴重阻礙復原進度。

雖然科學家已經發展出某些經驗法則來描述這些餘震的可能大小和發生時間，像是巴特定律(Bäth's Law)和大森定律(Ohmori's Law)，但要整理出可以預測餘震地點的法則卻是難上許多。

不過，地球和行星科學教授Brendan Meade以及在他實驗室工作的博士後研究員Phoebe DeVries受到Google研究人員提出的建議啟發，決定利用人工智慧技術來試著處理這項難題。

這對研究人員運用深度學習演算法(deep learning algorithm)來分析含有世界各處地震的資料庫，試著預測餘震可能發生的地點。他們也發展出一套可以預測餘震的系統，雖然還不精確，但已經比隨機分配好上許多。記述這項研究成果的論文8月30日發表於《自然》(Nature)期刊。

「關於地震有三件事情是我們想知道的――發生的時間、大小和地點。」Meade表示，「在此研究之前，我們已經有經驗法則可以讓我們預測餘震的發生時間和大小，而我們正在進行的研究便是想要得知第三項要素，也就是地點。」

「我很期待機器學習在處理這類問題時，未來發揮的潛力可以到什麼程度――這是一個值得追蹤的重要議題。」DeVries表示，「餘震預測是一個特別適合機器學習去挑戰的問題。因為餘震的行為模式牽涉到許到物理作用的影響，而機器學習極為擅長挑出各個因素彼此之間的關聯。我認為我們現在的成果才觸及問題的表層而已，對於餘震預報我們還有很多能做的……這真的相當令人興奮。」

Meade數年前待在劍橋的Google進行兩年的休假研究期間，首度有了試著利用人工智慧神經網路(neural network)來預測餘震的想法。

Meade說他在跟處理相關問題的研究團隊共事時，有位同事提出即將問世的深度學習演算法或許能讓這個問題變得更容易處理。Meade之後和DeVries一同針對餘震方面的問題進行研究。DeVries之前的研究是運用神經網路把高效能運算程式轉化成可以在筆記型電腦上執行的演算法。

Meade說：「(科學家的)終極目標是要做出完整的預報，而我們希望能有所貢獻。」

為了進行這項研究，Meade和DeVries首先取得了199個大地震之後的餘震觀測資料。

「在規模5或更大的地震發生之後，研究人員花了許多時間去繪製斷層的哪些部份發生滑動且滑動了多少。」Meade表示，「許多研究可能只會用上一或兩個地震之後的觀測結果，但我們的研究運用了整個資料庫……我們還結合了以物理原理為基礎、用來模擬地震之後大地受到的應力和應變如何變化的模型，以及主震造成的應力和應變可能是餘震形成原因的理論。」

在備有這些資訊之後，它們接著將地震發生的區域分割成以5公里為邊長的立體網格。系統會依次確認每個網格當中是否有餘震發生，然後命令神經網路尋找餘震發生地點和主震產生的應力之間的關係。

「問題在於哪些因子結合起來可能具有預測能力。」Meade表示，「之前有很多人提出相關理論，但這篇論文得到的結果之一確實推翻了現今最為主流的理論。我們得出該理論的預測能力相當微弱，反而是另一個理論的預測能力高出許多。」

Meade說系統指出預測力最佳的變量是「軸差應力張量之第二應力不變量(second invariant of the deviatoric stress tensor)」，更常用的簡寫為J₂。

「此變量通常是用在冶金學或其他理論中，在地震科學中從來沒有受到矚目。」Meade表示，「不過，這代表神經網路不會得出異想天開的結果，而是具有高度的可解釋性(interpretability)。此外，最棒的一點是我們還能找出科學家應該去關注哪些物理作用。」

DeVries說可解釋性相當重要，因為長久以來許多科學家把人工智慧系統視為黑盒子――可以憑藉某些資料就產生結果。

「這是我們研究過程裡最重要的一部份。」她說，「當我們剛開始訓練神經網路，我們發覺它在預測餘震發生位置表現得非常好，但我們認為更重要的是可否解讀出系統進行預測時，對它來說哪些是重要或者有用的因素。」

不過真實世界的數據是極端複雜的，要從中解讀系統如何預測會是相當艱鉅的任務。因此研究人員改要求系統對高度理想化的合成地震進行預測，然後再檢視預測成果。

「我們檢視神經網路輸出的預測結果，然後再檢視如果以不同變量為主來進行地震預測時應該得到的結果。」她說，「比較它們的空間分布，我們可以得出J₂似乎是系統進行預測時最重要的依據。」

Meade表示用來訓練神經網路的地震和餘震事件遍布全球各地，所以得出的系統對於許多種斷層類型產生的餘震都能適用。

「世界不同地方的斷層幾何型態也不同。」Meade表示，「在加州，多數斷層都是走向滑移斷層；但在其他地方，比方說日本，就有十分淺的隱沒帶。這個系統優秀的地方在於你可以用這些地方的資料訓練它之後，用它來預測其它地方的餘震，因此這個系統確實具有通用性。」

「要實際預測餘震我們還有很長的路要走。」她說，「雖然我們離做出即時預測的距離還相當遙遠，但我認為機器學習在這方面的潛力不可小覷。」

Meade表示他的下一步是嘗試用人工智慧技術預測這些地震的規模，目標是有朝一日可以幫助人類預防地震造成重大損害。

「正統的地震學家大部分像是病理學家，」Meade如此比喻，「他們在地震災害發生之後對其進行研究。但我想做的不是這些――我更想成為流行病學家――我想瞭解促成這些事件的條件、原因以及過程。」

最後，Meade表示這項研究提供了一個範例，顯示深度學習演算法具有潛力可以回答直到最近科學家連提問都相當困難的問題。

「我認為結果讓我們對地震預測的想法有了重大變革。」他說，「地震預測不再是完全遙不可及的概念。這項研究的結果不只是相當有趣而已，我認為它具體而微地呈現出在人工智慧時代，所有科學領域重建時大體上都會經歷的變革。」

他接著表示：「最近幾年某些令人望而生畏的難題變得相當容易處理，不只是因為電腦運算能力增加――科學界從此獲益良多，原因是……。AI這個字聽起來雖然相當嚇人，但實際上並非如此。它是一種十分平易近人的計算方法，我想現在有許多人越來越瞭解AI到底是什麼了。」

Earthquakes: Attacking aftershocks

Study uses AI technology to begin to predict locations of aftershocks

In the weeks and months following a major earthquake, the surrounding area is often wracked by powerful aftershocks that can leave an already damaged community reeling and significantly hamper recovery efforts.

While scientists have developed empirical laws, like Bäth's Law and Ohmori's Law, to describe the likely size and timing of those aftershocks, methods for forecasting their location have been harder to grasp.

But sparked by a suggestion from researchers at Google, Brendan Meade, a Professor of Earth and Planetary Sciences, and Phoebe DeVries, a post-doctoral fellow working in his lab, are using artificial intelligence technology to try to get a handle on the problem.

Using deep learning algorithms, the pair analyzed a database of earthquakes from around the world to try to predict where aftershocks might occur, and developed a system that, while still imprecise, was able to forecast aftershocks significantly better than random assignment. The work is described in an August 30 paper published in Nature.

"There are three things you want to know about earthquakes -- you want to know when they are going to occur, how big they're going to be and where they're going to be," Meade said. "Prior to this work we had empirical laws for when they would occur and how big they were going to be, and now we're working the third leg, where they might occur."

"I'm very excited for the potential for machine learning going forward with these kind of problems -- it's a very important problem to go after," DeVries said. "Aftershock forecasting in particular is a challenge that's well-suited to machine learning because there are so many physical phenomena that could influence aftershock behavior and machine learning is extremely good at teasing out those relationships. I think we've really just scratched the surface of what could be done with aftershock forecasting...and that's really exciting."

The notion of using artificial intelligent neural networks to try to predict aftershocks first came up several years ago, during the first of Meade's two sabbaticals at Google in Cambridge.

While working on a related problem with a team of researchers, Meade said, a colleague suggested that that the then-emerging "deep learning" algorithms might make the problem more tractable. Meade would later partner with DeVries, who had been using neural networks to transform high performance computing code into algorithms that could run on a laptop to focus on aftershocks.

"The goal is to complete the picture and we hope we've contributed to that," Meade said.

To do it, Meade and DeVries began by accessing a database of observations made following more than 199 major earthquakes.

"After earthquakes of magnitude 5 or larger, people spend a great deal of time mapping which part of the fault slipped and how much it moved," Meade said. "Many studies might use observations from one or two earthquakes, but we used the whole database...and we combined it with a physics-based model of how the Earth will be stressed and strained after the earthquake, with the idea being that the stresses and strains caused by the main shock may be what trigger the aftershocks."

Armed with that information, they then separate an area found the into 5-kilometer-square grids. In each grid, the system checks whether there was an aftershock, and asks the neural network to look for correlations between locations where aftershocks occurred and the stresses generated by the main earthquake.

"The question is what combination of factors might be predictive," Meade said. "There are many theories, but one thing this paper does is clearly upend the most dominant theory -- it shows it has negligible predictive power, and it instead comes up with one that has significantly better predictive power."

What the system pointed to, Meade said, is a quantity known as the second invariant of the deviatoric stress tensor -- better known simply as J₂.

"This is a quantity that occurs in metallurgy and other theories, but has never been popular in earthquake science," Meade said. "But what that means is the neural network didn't come up with something crazy, it came up with something that was highly interpretable. It was able to identify what physics we should be looking at, which is pretty cool."

That interpretability, DeVries said, is critical because artificial intelligence systems have long been viewed by many scientists as black boxes -- capable of producing an answer based on some data.

"This was one of the most important steps in our process," she said. "When we first trained the neural network, we noticed it did pretty well at predicting the locations of aftershocks, but we thought it would be important if we could interpret what factors it was finding were important or useful for that forecast."

Taking on such a challenge with highly complex real-world data, however, would be a daunting task, so the pair instead asked the system to create forecasts for synthetic, highly-idealized earthquakes and then examining the predictions.

"We looked at the output of the neural network and then we looked at what we would expect if different quantities controlled aftershock forecasting," she said. "By comparing them spatially, we were able to show that J₂ seems to be important in forecasting."

And because the network was trained using earthquakes and aftershocks from around the globe, Meade said, the resulting system worked for many different types of faults.

"Faults in different parts of the world have different geometry," Meade said. "In California, most are slip-faults, but in other places, like Japan, they have very shallow subduction zones. But what's cool about this system is you can train it on one, and it will predict on the other, so it's really generalizable."

"We're still a long way from actually being able to forecast them," she said. "We're a very long way from doing it in any real-time sense, but I think machine learning has huge potential here."

Going forward, Meade said, he is working on efforts to predict the magnitude of earthquakes themselves using artificial intelligence technology with the goal of one day helping to prevent the devastating impacts of the disasters.

"Orthodox seismologists are largely pathologists," Meade said. "They study what happens after the catastrophic event. I don't want to do that -- I want to be an epidemiologist. I want to understand the triggers, causing and transfers that lead to these events."

Ultimately, Meade said, the study serves to highlight the potential for deep learning algorithms to answer questions that -- until recently -- scientists barely knew how to ask.

"I think there's a quiet revolution in thinking about earthquake prediction," he said. "It's not an idea that's totally out there anymore. And while this result is interesting, I think this is part of a revolution in general about rebuilding all of science in the artificial intelligence era.

"Problems that are dauntingly hard are extremely accessible these days," he continued. "That's not just due to computing power -- the scientific community is going to benefit tremendously from this because...AI sounds extremely daunting, but it's actually not. It's an extraordinarily democratizing type of computing, and I think a lot of people are beginning to get that."

原始論文：Phoebe M. R. DeVries, Fernanda Viégas, Martin Wattenberg, Brendan J. Meade. Deep learning of aftershock patterns following large earthquakes. Nature, 2018; 560 (7720): 632 DOI: 10.1038/s41586-018-0438-y

引用自：Harvard University. "Earthquakes: Attacking aftershocks: Study uses AI technology to begin to predict locations of aftershocks." ScienceDaily. ScienceDaily, 29 August 2018.