A hymn to entropy (Invited talk)

Koutsoyiannis, D

dc.contributor.author	Koutsoyiannis, D	en
dc.date.accessioned	2014-03-01T02:54:05Z
dc.date.available	2014-03-01T02:54:05Z
dc.date.issued	2011	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/36590
dc.title	A hymn to entropy (Invited talk)	en
heal.type	conferenceItem	en
heal.publicationDate	2011	en
heal.abstract	While entropy, and in particular its tendency to become maximum, is typically regarded as a curse, I contend that it is an eulogia. Not only does it offer the basis to understand and describe Nature, but it also constitutes the driving force of change and evolution. Entropy is a measure of uncertainty, defined within probability theory, and its maximization offers a powerful principle, applicable to both description of physical systems and logical inference. In thermodynamics dealing with the equilibrium state of systems with hugely many components (molecules) identical to each other or belonging to a few kinds, application of the principle of maximum entropy results in macroscopic certainties. These are, au fond, statistical laws based on maximization of uncertainty at the microscopic level, yet yielding extremely low macroscopic uncertainty, so low that we often misinterpret the laws as deterministic. However, the formation of clouds and the precipitation cannot be described in terms of systems with identical elements. Furthermore, the flow of water on Earth and its spatial and temporal variability are even more difficult to model because the relevant systems (catchments, rivers, aquifers) are composed of extremely diverse elements. In the last decades, the dominant target and aspiration in hydrological sciences has been the radical reduction of uncertainty. I contend that this aspiration traces a research direction that is wrong and opposite to how Nature works. In contrast, a promising path to faithfully model hydrological processes and systems should be sought in extremization of entropy (i.e. uncertainty).	en