<<

NEWS & VIEWS |Vol 453|15 May 2008

OBITUARY Edward N. Lorenz (1917–2008) Meteorologist and father of .

Edward Norton Lorenz, whose pioneering nonlinear such as the atmosphere. studies of atmospheric dynamics led to Most immediately for Lorenz’s , MIT his accidental discovery of , this meant that long-term died of cancer at his home in Cambridge, predictions were impossible, because the Massachusetts, on 16 April. A modest, atmosphere’s initial state can never be unassuming and kind man, his personal specified precisely enough. That was a qualities and intellectual insights had situation that increased computing power been a constant feature in the field of could not change. for more than 60 years; he Lorenz perfectly encapsulated this co-authored his last paper just weeks before unknowability in the title of a talk that his death. he gave to the American Association for Born on 23 in West Hartford, the Advancement of Science in 1972. The Connecticut, Lorenz took bachelor’s question it asks has since lodged itself in the and master’s degrees in at public’s consciousness: “Does the flap of a , New Hampshire, and butterfly’s wings in Brazil set off a tornado , respectively. Service in Texas?” But the influence of chaos theory as a weather forecaster for the US Army extends far beyond meteorology, and much Air Corps during the Second World War deeper: it challenges the entire deterministic led him into meteorology, and he received a world view, as was confidently expressed, doctorate in the subject at the Massachusetts for instance, by the mathematician and Institute of Technology (MIT) in 1948. philosopher Pierre-Simon Laplace, who He remained in MIT’s Department of stated at the beginning of the nineteenth Meteorology for the rest of his academic century that the entire could be career, becoming emeritus professor determined by constructing and solving there in 1987. the governing all components of Lorenz made crucial contributions to at a consistent rate. When Lorenz plotted the Universe. atmospheric science, many of which are variables representing temperature and Although the existence of chaos had been still routinely taught to students and widely flow against one another, the recognized before Lorenz — notably in the used in weather . Perhaps eventually adopted that 1890s by Henri Poincaré, in his study of the foremost among these is his formulation in traced out something akin to a pair of motions of three or more gravitating celestial the mid-1950s of the concept of ‘available butterfly wings — a since called bodies — it was Lorenz’s meteorological potential energy’, which he used to explain the Lorenz . He further observed that demonstration and analysis that established how potential energy and kinetic energy the system moved from one wing the universal applicability of the concept, and are interchanged in the atmosphere. His of the butterfly to another in a seemingly earned him the title ‘the father of chaos’. But application of these ideas culminated in erratic manner. it took a decade for chaos theory to percolate his influential book of 1967, The Nature In his book The Essence of Chaos, Lorenz through to the general scientific community. and Theory of the General Circulation of the recounts how he came to discover the When it finally did, it launched a revolution, Atmosphere. He was also instrumental in extreme sensitivity of his model to small rapidly extending its sway into many fields of the development of numerical techniques changes. Wishing to repeat his simulation, , chemistry, and for weather prediction. One example he restarted it with that had been — and, in doing so, becoming part of the — again, still widely used — is his scheme printed out for the start conditions, and popular lexicon. for the numerical treatment of changes in left it to go down the hall to fetch a cup of Lorenz received many honours and prizes atmospheric variables with height, now coffee. On his return, he found that the in recognition of his work, among them the known as the Lorenz vertical grid. result was nothing like the previous one. — established by the Royal But the work for which Lorenz is He soon identified the reason: the numbers Swedish Academy of Sciences to recognize undoubtedly most widely known is a from the print-out were rounded off. In the work in fields not covered by the Nobel now-classic paper published in the Journal course of a coffee break, that small error had prizes — in 1983, and the Kyoto Prize in of Atmospheric Science in 1963. Entitled propagated with exponential speed to change 1991. The citation for that prize lauded “his ‘Deterministic nonperiodic flow’, it presented the result completely. boldest scientific achievement in discovering surprising results from a simplified This discovery was epoch-making for two ‘deterministic chaos’, a principle that has computational model that simulated thermal reasons. The first lay in Lorenz’s integration profoundly influenced a wide range of basic convection in a fluid layer heated from below of analytical methods with computational sciences and brought about one of the most and cooled from the top. The calculated simulations, with which he — albeit with a dramatic changes in mankind’s view of nature flow of the fluid was extremely irregular, pre-1960 computer that was bulkier, noisier since Sir Isaac Newton”. with almost random qualities. But more and vastly slower than the PCs of today — set importantly, it exhibited extremely sensitive an early precedent for a mode of research Edward Ott is in the Departments of Physics, dependence on initial conditions: two fluid that has since become a norm. But much and of Electrical and Computer Engineering, states that were at first just slightly different more profound ramifications stemmed from University of Maryland, College Park, diverged from each other exponentially, Lorenz’s realization of just how general the Maryland 20742, USA. with their differences doubling repeatedly types of motion he had uncovered were in e-mail: [email protected]

300