Но если читаете по-английски, то вот (из третьей части):

"Let us try to analyze behavior of mountain glaciers from the point of view of systems approach in more details, and look for manifestations of systemic relations in glaciers, which could be easily identified and used for constructing a cybernetic model of glacier system.

Glacier behavior is usually considered as an indicator of environmental conditions. However, even from the first sight it is clear that changes of the same environmental parameters should be reflected in glacier behavior differently if one considers them at different spatio-temporal scales.

Glacier behavior can be expressed in various terms. For instance, one can consider fluctuations of glacier shape, like advance and retreat of terminus, change of surface area, fluctuations of surface topography and elevation, etc., as one of the aspects of its behavior. Another aspect is fluctuations of some dynamical characteristics, like ice flow velocity, shear stress and so on. In order to be able to compare glacier behavior at different spatio-temporal scales, one has to select such a term from this list, which could be clearly distinguished and easily detected at the most of possible scales. Apparently, ice flow velocity is one of such terms.
Let us consider changes of ice flow velocity under the environmental impact. Increase of ice melting within particular short period provokes acceleration of ice flow within some parts of glacier tongue due to additional lubrication at the bottom. The same increase of melting considered at the scale of several years leads to thinning of glacier and, consequently, to fall in general trend of glacier’s mean ice flow velocity. It is obvious, that in these cases one reason leads to different, even opposite consequences.

This simple example shows that different laws govern glacier behavior at different spatio-temporal scales. Generally speaking, this is trivial for macroscopic systems consisting of many components or subsystems. A macroscopic body consisting of multitude of atoms appears and behaves in a different way than an atom or elementary particles an atom consists of. However, in physics it is proved that the laws of Newtonian mechanics present a particular, extreme case of the laws of quantum mechanics.

At the same time, in glacier science there is still a lack of knowledge about how different relational (spatio-temporal, hierarchical) levels correspond to each other. Moreover, although that qualitative gap between those levels obviously exists, many researchers try to extrapolate laws of small spatio-temporal scale to the processes significantly elongated in time and related to the whole glacier. When modeling long-term reaction of the whole glacier to climatic disturbances, they build their models on the base of flow law obtained in short-term experiments with rather small pieces of ice (Paterson, 1994). This seems methodologically incorrect because we still do not know how the flow law could transform due to transition from small spatio-temporal scale to the large one.

The question is: can be the laws governing glacier behavior at the large spatio-temporal scale educed as a particular case of the laws governing its behavior at the small spatio-temporal scale, as it was demonstrated by Ehrenfest for Newtonian laws and quantum mechanics, respectively? There are two possibilities. If the answer is “yes” (which is usually characteristic for rather simple systems), then reductionistic analysis is all one needs when modeling a glacier. At the large spatio-temporal scale, glacier behavior will appear as “summation” of behavior of its components at the low scale, as a kind of “average” behavior defined by summary action of all those factors governing the low scale changes.

On the contrary, if the answer is “no”, then one deals with a phenomenon that is quite different from the things considered by traditional reductionism – the phenomenon of self-organization. In this case glacier should be considered as a system that is more than just a sum of its parts. Such a system is organized in a special way so that its components functioning at the low level could produce a necessary effect at the level of the whole system. Properties of their functioning are caused by the demands of the system’s organization and guaranteed by the mechanisms of downward causation.

Science, and physics in particular, has developed out of the Newtonian paradigm of mechanics. In this world view, every phenomenon we observe can be reduced to a collection of atoms or particles, whose movement is governed by the deterministic laws of nature (Heilighen, 2001). That is the first explanation of those two stated above. As in glaciology the problem of relation between different spatio-temporal scales still was not considered thoroughly enough, the first explanation is assumed as default, because it lies in the basement of the classical science.

However, the twentieth century science, as it was described in II.4, has slowly come to the conclusion that such a philosophy is not the only possible point of view. Around the middle of the century, researchers from different backgrounds and disciplines started to study phenomena that seemed to be governed by inherent creativity, by the spontaneous appearance of novel structures or the autonomous adaptation to a changing environment. The different observations they made, and the concepts, methods and principles they developed, have slowly started to coalesce into a new approach, a science of self-organization and adaptation (Heilighen, 2001). After science had faced this previously not considered phenomenon, the numerous examples demonstrated that it was not something unusual, but, on the contrary, something immanent for the most of macroscopic systems.
The correlation or coherence between separate components of a system, produced by self-organization defines an ordered configuration. Organization can also be defined thus as the characteristic of being ordered or structured so as to fulfil a particular function. In self-organizing systems, this function is the maintenance of a particular configuration, in spite of disturbances (homeostasis). Only those orders will result from self-organization that can maintain themselves (Rosnay, 1997; Heilighen, 2001).

Self-organizing systems are thermodynamically open, but organizationally closed. For the outside observer, closure determines a clear distinction between inside (the components that participate in the closure) and outside (those that do not), and therefore a boundary separating system from environment. Organizational closure turns a collection of interacting elements into an individual, coherent whole. Its higher level, emergent property constrains the behavior of the lower level components by downward causation (Heilighen, 2001).

The goal of the present paragraph (III.1), thus, is to understand which one of these two described causations we deal with in case of mountain glacier, and to clarify relationship between micro- and macro-levels of glacier systems, using information about variations of their behavior at different spatio-temporal scales. In further analysis, observational data from several temperate glaciers from different mountain regions of the world will be used, where measurements of ice flow velocities on small or large spatio-temporal scales had taken place.

[далее идут данные полевых наблюдений на разных ледниках, с подробным описанием, где, когда, кем и что измерялось - сначала для небольших пространственно-временных шкал, затем - для больших; думаю, приводить их все здесь особенно не зачем, остановимся на выводах]

From the foregoing consideration, it is clear that mountain glacier behavior significantly differs for small and large spatio-temporal scales. In the first case, when one considers changes of a small part of glacier for a short period of hours, days or weeks, glacier behaves as a near chaotic, unpredictable medium, governed with numerous factors difficult to be taken into account properly. However, in the second case, when the whole glacier is observed for a long period of number of years, it demonstrates steadiness and good predictablilty, being governed by glacier mass generally. Thus, glacier behavior at different scales is governed by fundamentally different laws.

If one makes an attempt to understand how particular chaotic behavior generates well-ordered one for glacier system as a whole, the main question can be formulated as it was done in the beginning of this paragraph: can be the laws governing glacier behavior at the large spatio-temporal scale educed from the laws governing its behavior at the small spatio-temporal scale? If the answer is “yes”, then at the large spatio-temporal scale, glacier behavior will appear as “summation” of behavior of its components at the low scale. On the contrary, if the answer is “no”, then one deals with self-organization and downward causation, and reductionistic analysis is unjustified. In this case glacier should be considered as a system that is more than just a sum of its parts.

In the context of the brief excursus into science of self-organization (synergetics) in the beginning of III.1, it becomes essentially clear that synergetics studies exactly the same thing one can discern in case of glaciers when analyzing it systematically and impartially. If one considers behavior of glaciers at different spatio-temporal scales from the point of view of synergetics, some obvious conclusions could be made.

First of all, from this viewpoint, it is possible to decide how transition from low level of relations to the higher one is realized within a glacier. Since an assumption of classic science that multiple chaotic fluctuations could generate order regularly on their own is quite unrealistic, the only reasonable explanation for the transition is that glacier as a system controls general trends of those chaotic fluctuations with some systemic mechanisms.

That is a manifestation of downward causation.

Glacier acts as an information processing machine selecting such fluctuations of low level which finally could – due to their collective or synergetic action – allow realization of glacier at the higher level as a system governed by its own systemic properties, that is to say, by the laws of functioning of the glacier system as a whole. At this level, according to the previous analysis, one of the order parameters of glacier system is total glacier mass, which can be characterized as “leading slow variable” effective at the timescale of years.

Characteristic variables of smaller spatio-temporal scales are forced to adjust themselves to this order parameter, as it is known from the theory of inertial manifolds (Kapitsa et al., 2001)."