Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://chronos.msu.ru/old/EREPORTS/korotaev_experimental_advanced.pdf Äàòà èçìåíåíèÿ: Sat Dec 14 12:32:16 2013 Äàòà èíäåêñèðîâàíèÿ: Fri Feb 28 20:43:52 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: cosmic rays

© Korotaev S.M.

Experimental Study of Advanced Correlation of Some Geophysical and Astrophysical Processes
S.M. Korotaev Geoelectro magnet ic Research Center of Inst itute of Phys ics of the Earth, Russian Academy of Sciences (GEMRC P.O. Box 30 Troitsk, Moscow Region 142190 Russia). Fax +7-495-7777218, e-mail serdyuk@izmiran.rssi.ru Abstract Macroscopic nonlocalit y represents itself in correlation of different dissipat ive pro cesses without any local carriers o f interact ion and with Bell-t ype inequalit y vio lat ion. Nonlocal correlat ion obeys weak causalit y principle. It invo lves the possibilit y of advanced transaction between the random dissipative pro cesses. Wide series o f lo ng-term exper iments on observation of correlation of insulated lab probe-processes with the large-scale source-processes have been performed. For the natural random source-processes: the solar, meteorological and geo magnet ic act ivit y the advanced react ion was reliable detected. Moreover advanced correlation pro ved to be stronger than retarded one. Due to high level o f advanced correlat ion forecast ing applicat ions have a sense. This possibilit y was demonstrated by the forecast of the random component of geomagnetic activit y. Keywords: nonlocalit y, dissipation, causalit y, entropy, forecast

1

Introduction

Pheno menon o f quantum nonlocalit y at tracts increasing attention due t o number o f its unusual properties. In part icular, t ransactional interpretation o f quantum nonlocalit y in the framework of Wheeler-Feynman action-at-a-distance electrodynamics [2] suggests existence of signals in reverse time. According to principle o f weak causalit y [2] it leads to observable advanced correlat ion o f unknown (not determined by evolut ion) states [3] or, in other terms, rando m processes. Further, although it is generally supposed that nonlocalit y exists only at the micro-level, there is reason to believe that it asymptotically persists in the strong macro-limit and it has been proved in the numerica l [5] and real [4, 7] experiments. Moreo ver a new way o f entanglement format ion via a commo n thermostat (which can be served the electromagnet ic field) has been suggested [1] and this way needs dissipat ivit y o f the quantum-correlated processes (namely radiation ones). It means that dissipat ivit y may not only lead to decoherence, but on the contrary, it may play a constructive role. On the ot her hand, more than 30 years ago N.A. Kozyrev had suggested causal mechanics theory and conducted the various experiments [19], which originated from the idea o f fundamental time asymmet ry, but led to macroscopic pheno mena similar to microscopic nonlo cal ones. Specifically, he had observed correlat ion o f the probe dissipa-


tive pro cess (in the telescope detector) with large-scale ones of the stars with three t ime shifts, corresponding classical retardation, symmetrical advancement and zero between them, i.e. instantaneous [20]. According to causal mechanics such correlation of differ-


ent dissipat ive processes was explained by not any local carr iers of interact ion, but by some physical pro perties of t ime as an act ive substance. Kozyrev's theoret ical and experimental conclusio ns were so unexpected (with weakly formalized theory and not too strict performance of the experiment), that they could not be accepted in due course. But in 1990 s similarit y o f the results o f causal mechanics and so me recent ones of quantum mechanics had beco me obvious. Understanding of causal mechanics effects as possible manifestation o f quantum nonlocalit y at t he macro-level had allo wed to perform the experiments showed availabilit y o f advanced correlation [11 - 18]. In this art icle obtained result s are generalized and particular attention to the most recent ones is paid. In Sec. 2 theoretical ideas and in Sec. 3 experimental ones are presented. In Sec.410 experimental data, results of their processing and interpretation are shortly described. Conclusio n is in Sec. 11.

2 Heuristic Model
As a development of strict theory of macro scopic nonlocalit y is very difficult problem, we consider the simplest heurist ic model. We fo llow Cramer's interpretation o f quantum nonlocalit y within Wheeler-Feynman action-at-a-distance electrodynamics [2], but we use the latter in modern quantum treatment [6]. This theory considers direct particle field as superposit ion of retarded and advanced ones. The advanced field is unobservable and manifests only via radiation damping, which is the dissipative process. But first of all let's take notice to likeness of axioms of causal mechanics and actionat-a-distance electrodynamics. In the electrodynamics transact ion of the charges separat ed by finite distance 8x and lapse 8t (with zero interval) is postulated. Self-act ion o f the charges is absent. Two from three Kozyrev's axio ms [19] (there are 8x^0 and 8t 0 between any cause and effect) assert the same, replacing only terms «charges» by «cause» and «effect» (third axio m postulates time asymmetry), w hile in Ref[20] N.A. Kozyrev grounded that transact ion occurs through zero interval. In Ref[8] uncertaint y of the terms «cause» and «effect» had been removed. Essentialit y o f the formalism is as follo ws. For any observables X and Y t he ind ependence funct ions / can be introduced:
S(Y|X) () S(X |Y) 0^ ^1 () (1)

where S denote condit ional and marginal Shannon ent ropies. For example if Y is singlevalued function of X then iYX = 0, if Y does not depend on X, then iYX = 1. Roughly saying, the independence funct ions behave inversely to module o f correlation one. Next, the causalit y funct ion is considered: y=
X\Y

,

0
(2)


It can define that X is cause and Y is effect if <1. And inversely: Y is cause and X is effect if y>1. The case y=1 means non-causal re lat ion X and Y (t hey are related with some commo n cause).Theoretical and mult iple o f exper imental examples have shown that such formal definit ion of causalit y does not contradict its intuit ive understanding in obvious situat ion and can be used in non-obvious ones (e.g.[8 - 10]). Our definit io n allows for mulating all three Kozyrev's axio ms in the for m of one: y<1^>tY >tx,xr ; y>1^>tY xx.

(3)

Stat ement (3) is very natural, but it is axio m o f strong or local causalit y. For nonloca l transaction this stat ement might be invalid due to advanced correlat ions of the dissipative processes. I ndeed any dissipat ive process [21 ] is ult imately related to the radiat ion and therefore to the radiation damping. As a result it can be shown that advanced field connects the dissipat ive processes [12]. Time asymmetry is expressed as absorption asymmetry. While absorption of retarded field is perfect, absorpt ion of advanced one must be imperfect. Having accepted that total field E is superposit ion E = AE + BE , (4) where A and are co nstants, and having denoted efficiency o f absorption o f retarded field by a (a = 1 corresponds to perfect absorption, a = 0 - to absence o f absorption), advanced one by b, it is easily to obtain [6], that 1-b 1- A =----------- , B = ----------- , 2--a--b 2--a--b (5)

Subst itution to Eq. (4) A = 1, = 0 corresponds to really o bserving situat ion, t hat is compatible with Eq. (5) only if a = 1, 0 < # < 1. It should be stressed wide a priori arbitrariness in value b, which may be clo se as to unit so to zero. Therefore the screening pro perties of the matter must be in one degree or another att enuated. The fact itself o f imperfect absorption o f the advanced field means a possibilit y o f its separate detection. From the operational considerat ion it is possible to formulate the following equat ion:
X S = 2 ,-

(6)

s
2 2

v

v

JX


where
4 2 2 <5 ------- 24 , v < ,

mee

(7)


S is the entropy production in a probe process (that is detector), s is densit y of entropy production in the sources,
3

Experimental Technique

The task of the experiment is to detect the entropy change of the environment according to Eq. (6) under condition that all known kinds o f classical local interaction are suppressed. Although any dissipative process may be used as the probe one, its choice is dictated by relat ive value o f effect and theoretical dist inctness, allowing to relate the measured macro-parameter (signal) with the left-hand side of Eq. (6) and consciously to take steps on screening and/or control of all possible local no ise-factors. Two experimental setups for study o f macroscopic nonlo calit y had been constructed [11, 14]. In the Geoelectromagnetic Research Inst itute (GEMRI) setup two types o f detectors based on spontaneous potential variations o f weakly po larized electro des in an electrolyte and on spontaneous variat ions of the dark current of the photomult iplier had been used. In the Center of Applied Physics (CAP) setup ion mo bilit y detector based on spontaneous variat ions of conduct ivit y fluctuation dispersio n in a small electrolyte volume had been used. As in this paper the results mainly wit h GEMRI setup are considered, we concern only its detectors. Theory of detecto rs [11,12,18] allows to relate the measured signals with the rate of entropy production in the probe-process. Final formulae in small amplitudes appro ximat ion are:
1 |4 | =--1=-------- AU , (8)

6
AS =--------- , (9)

where q is io n charge, is t emperature, U is measurable variable electrode potentia l difference, /is measurable photomult iplier dark current. All known local factors influencing on U: temperature, pressure, chemism, illumination, electric fie ld, concentration and mo vement of the electrolyte must be excluded. Analogously, no ise-forming facto rs influencing on / to be excluded are: t emperature, electric and magnetic field, illumination, mo isture and feed vo ltage instabilit y. Design of the detectors provides this exclusio n. All technical det ails about design o f the detec-


tors and their parameters are presented in Ref [11, 14, 17, 18].


The GEMRI setup consists of nearby electrode and photomult iplier detectors, another electrode detector spaced at 300 m and apparatus for the local factors control. The CAP setup with ion mobilit y detector operating under well controlled local condit ions is spaced at 40 km from GEMRI one. It is known that quantu m nonlocalit y vio lates strong causalit y and persists weak one [2]. It means, that if a source process is noncontrolled (random), we ca n observe both retarded and advanced correlat ions. But if an observer init iates a source-pro cess, only retarded correlat ion is possible. That is why the most interesting source process are random large-scale natural ones. The experiment described below was devoted to study detectors reaction on various geophysical and astrophysical processes with big rando m co mponent. The exper iments wit h controlled lab artificial source-processes had also been conducted, though they had, of course, demonstrated only retarded correlat ion.

4

Data and Pr ocessing

The experiments with natural source processes were long-term (with duration of continuous series not less than several months). They were conducted in 1993-96 wit h the electrode detector, in 1996-97 with the all 3 detectors of the GEMRI setup and in 1997 with CAP setup, and in 2001-2003 again wit h GEMRI and CAP setups. Except t he detector signals the fo llowing parameters were measured: internal detecto rs temperature (residual variat ions strongly suppressed by thermostating system) accurate to 0.001 K, external (lab) t emperat ure ­ 0.1 K, outdoor (atmospheric) t emperature ­ 0.1 K and geomagnet ic field ­ 0.01 nT. Sampling rate was chosen fro m 1m to 1h. In addit ion hourly data on cosmic ray counting rat e (as one more reasonable no ise-factor) and atmospheric pressure ( as index of nonlo cal influence of syno ptic act ivit y) were taken fro m nearby IZMIRAN neutron monitor. Standard international data on the g lo bal geo magnetic (Dst and Ap indices) and so lar (radio wave flux at 9 standard frequencies wit hin range 245 ... 154000 MHz and also X-ray flux fro m GOES satellite) act ivit y were taken to study the most large-scale processes. Data were processed by the methods of causal, correlational, regressional and spectral analysis. Algorithm of the causal analysis was described in Ref. [8 ­ 10, 14,18] in detail. The main point is calculation of conditional and marginal probability distributions of detector signals (X) and source processes indices (Y) by the time series. The Y series where taken with enough long "tails" before and after the X series ends to provide calculation of the distributions and their entropies in corresponding time shift range. Other processing methods were standard.

5

Relation of the Signals of Different Detectors

So we had long-term measurements with 4 detectors of 3 t ypes. Their signals proved to be rather high and synchronous correlated. For any pairs maximum of correlat io n funct ion r achieves 0.7 ­ 0.8 at time shift t= 0 . Mathemat ical exclusio n o f single possible commo n local factor not completely suppressed by screening, namely interna l temperature (other non-suppressed local factors ­ the magnet ic field and cosmic rays


proved to be not influencing on the detectors within their sensit ivit y at all) leads to correlat ion increasing. Therefore there is not any local co mmo n cause of the signals. Level of correlat ion proved to be independent on type of detectors and on their separat ion within 40 km. Such correlat ion can be explained only in Cramer's sp irit [2]: by exchange of the detectors and some large-scale commo n sources (geophysical or astrophysical pro cesses) by the pairs of retarded and advanced signals, that is to be nonlocal. Due to that correlat ion we shall consider in the following sections mainly results wit h the electrode detector, which proved to be the most reliable and with which historically the greatest volume o f data was obtained.

6 Correlation of the Detector Signal with the Meteorological Activity
First of all temperature variations of the environment lead to its entropy changes. The problem is co mplicat ed by trivial local influence of small residual variat ions o f the internal temperature on the probe process, evoking weak retarded correlation. Thus for the electrode detector retarded correlat ion equals - 0.33 ± 0.02 at = -20.4 . But in the advanced domain, where correlat ion must be classically damped out, there is unusua l big correlat ion maximum 0.87 ± 0.01 at = 12.8 . Just at the same time shifts there are two minima of the independence functions. The advanced minimum is much deeper. Analysis of connection between the detector signal and external lab temperature has shown three maxima of correlation (minima of independence) at shifts 0 and ± 27 . The advanced minimum proved to be deepest and therefore could not be explained by any periodic effect. It corresponds to t heoretical prediction: we observe symmetrical retardation and advancement, the advanced signal is st ronger due to less abso rption by the intermediate medium. Availabilit y of the apparent synchronous signal can be explained by interference of the ret arded and advanced signals. Analysis o f influence o f the synopt ic act ivit y has also shown prevalence advanced correlat ion over retarded one. Level o f correlat ion and value o f advancement proved to be direct ly related wit h space scale o f the process. Thus as qualit ative index o f synoptic activit y simply at mospheric temperature can be taken (t ypical space scale is a few hundred km). In this case symmetrical by time shift advanced and retarded correlations have been revealed and level of advanced correlations are about twice as much ret arded ones. Maximum o f correlat ion (0.73 ± 0.01) is observed at advancement 13 days (Fig. 1). If the atmospheric pressure is taken as the index (space scale is a few t housand km) correlation pike achieves - 0.86 ± 0.01 at advancement 73 days. The same results are in terms of the independence and causalit y funct ion. The latter achieves 2.3, t hat means synoptic activit y is a cause o f the detector signal but the progress in reverse t ime! This result is independent on type o f detector [11].


Figure 1. Correlat ion funct ion o f the detector signal U and at mospheric temperat ure Ta rUTa. The t is time shift of Ta relative to U in days (negative t corresponds to retardation of U relative to Ta , positive t ­ to advancement).

7

Correlation of the Detector Signal with the Solar and Geomagnetic Activity

The solar activit y pro ved to be the most powerful dissipat ive process acting on detectors. It should be stressed that detectors are insensit ive to the so lar radio waves, their flux is only index o f the source entropy production.The detector signals proved to be most correlated with the solar radio wave flux in the frequency range 610...2800 MHz, corresponding to emissio n from the upper chromosphere ­ lower corona level, that is just fro m the level o f maximal dissipat ion the magneto-sound waves energy. The optimal frequency may change wit hin the ment ioned range in different years. The process of geo magnet ic act ivit y is an effect of the so lar one and it is weaker, but convenient for quant itative interpretation (Sec. 9). The variable magnet ic field is index of electric current dissipat ion in its source, that is magneto sphere (while the detectors are insensit ive to the local variable geo magnet ic field). The amplitude spectra of so lar radio wave flux R at the optimal frequency 610 MHz (R610), Dst-index of geo magnet ic act ivit y and detector signal U are shown in Fig. 2. All the spectra have two main maxima at period of solar rotation and its second harmonic. The spectrum o f the detector signal is obviously


more similar to the solar than to geomagnetic activit y. It is a consequence of direct influence of the so lar act ivit y on the probe process.

Figure 2. Amplitude spectra of the solar activity R610, geomagnetic one Dst and U in the period range T from 10 days to 274 days. For the analysis o f the ant icipatory effects the periodic co mponents were suppressed by filtration and we consider further only the random component. The qualitative resu lts are the same as in Sec. 6: advanced correlations exceed retarded ones and level of correlat ion increases along source space scale. Thus for magnetic field measured by setup's magnetometer advancement equals 2 days [11, 14, 17, 18], while for Dst-index o f global geo magnet ic act ivit y, reflect ing the most large-scale magnetosphere current systems it equals about month (it is not stable value; for different realizat ions and for different period range posit ion tof the main pike of g, i or r may be from 8d to 140d [11, 12, 15, 16]). Value of maxima l, i.e. advanced, g does not exceed 1.15 ( expectation errors of i and g are about 1%). The level o f advanced correlation wit h Dst after appropriate filtrat ion, increasing


signal/no ise ra-


tio (the no ise includes direct influence o f the Sun on the detector signal), can achieve 0.70­ 0.95. Herewith correlation t ime asymmetry (defined as max |r adv| / max |r ret|) in the shift trange ± 371d are wit hin from 1.10 ± 0.01 to 2.64 ±0.01 [12, 15]. Typical example o f correlat ion function showing advanced detector response at t = 42d is presented in Fig. 3.

Figure 3. Correlat ion funct ion of the detector signal U and geo magnet ic act ivit y Dst by data filtered in period range 364 > T > 28 days. The results o f causal analysis o f the detector signals and so lar act ivit y R have shown that in the advanced domain (t >0) values of the independence function are much less and ones of causalit y funct ion are much more than in the retarded domain (t <0), position tof the main pike of g, i or r may be fro m 42d to 280d. Value of maximal, i.e. advanced, gamounts up to 1.58, while r ranges into 0.50 ­ 0.92 (and relation wit h R is exp licit ly nonlinear). Big t- interval co rresponding to significant g >1 is explained by big volume o f the so lar at mosphere occupied by the source processes wit h diffusio n propagation [16]. An example of correlat ion functio n, corresponding advanced detector response at t = 42d is presented in Fig. 4. This is result of computation by the same t ime ser ies o f U as for Fig. 3 (without suppressing the annual period which is associated wit h the determined co mponent for Dst). For the given t ime series ret ardation o f Dst relat ive to R610 turned out small in our scale, therefore t = 42d in the both cases.


Figure 4. Correlat ion funct ion rUR of the detector signal U and so lar act ivit y R low-pass filtered data T > 28 days.

610

by

8

Bell-type Inequality Violation

Suppo se a process Z can influence on X only through Y alo ng the local causal chain: Z --> Y --> X. There is Bell-type sufficient co ndit ion of locality of Z - X connection [11, 14, 15, 18,]: /x|z>max(/x|7,/7|z). (10)

Sense of Ineq. (10) is simple: connect ion between the origin and end o f the chain can not be stronger t han in the weakest of two intermediate links. In our case Z and Y are some source processes, while X is a pro be process measured by the detector signal. We had opportunit y to t est Ineq. (10) in such a way that connect ion in Z - Y was known to be local (and certainly only retarded). In the first variant we used the rando m temperature variations (of external origin as there were no any heat sources inside of the detector dewar. Thus Z was external (lab) temperature and Y was internal one. For the advanced connections Z with X and Y Ineq. (10) was reliably vio lated, for symmetrical retarded ones was not (due to classical interaction) [11, 14, 18]. In the second variant we used the random variations of so lar (Z = R) and geo magnetic (Y = Dsi) activit ies. Retarded connection with X in this case was insignificant. For corresponding advanced connect ions Ineq. (10) was also reliably vio lated [15].


Consider this matter again by the most recent experiment, namely data of continuous measurements with the electrode detector of GEMRI setup. As compared wit h the previous experiments, the system o f its temperature stabilization was improved and thus signal/no ise rat io was magnified. Durat ion o f t ime series was 1 year (10/19/2002 10/18/2003). The detector signal (pot ential difference) f/was measured accurate to 0.5 JUV with data sampling 1 hour. As solar activity data we took daily solar radio flux R at optimal for the given case frequency 1415 MHz and two adjacent ones: 610 and 2800 MHz. Time series was taken for about 3 years (beginning 371 days before and finishing 371 days after the ends of U series). As geo magnetic activit y data we took internationa l hourly Dst-index for the same time as R. For co rrelat ion wit h R, U, and Dst data were previously daily averaged. We have been considering problem of detection of advanced correlation in more distilled performance (so did it in Ref [18, 19]). The matter of fact is, advanced correlat ion is physical property only the rando m processes. I f the determined, that is in given case periodic, components of variat ions are not suppressed, then advanced cross-correlat ion could be amplified by auto-correlat ion. It would be useful in forecasting pract ice, but here we are go ing to investigate namely advanced cro ss-correlat ion. Therefore we have to suppress the periodic co mponents with care. The main periodicit y in R (having a response in ¸/[16]) is synodic so lar rotation period. In addit io n, a lot o f geophysical processes have annual per iod, including its second harmonic. For these reasons U and R data were wide-band filtered in the per iod range 183 >i>28 . (For Dst because o f splitting of the spectral line corresponding to the solar rotation period, optimal lower bound of
_ d

the wide-band filtrat ion was more: 32 [12]). After this filtration the correlation function rUR was calculated in the time shift range 371 (T<0 corresponds to ret arded correlat ion r
_ | adv ret

, > 0 - advanced one r
d

).

Correlat ion time asymmetr y max rUR | / max |rUR |= 1.18 ± 0.06 , that is quite reliable.
adv

Maximal correlat ion rUR = 0.92 ± 0.03 is at advancement = 130 . At the adjacent frequencies the main maximum is also at = 130 , but level o f co rrelation is slight ly less: for 610 MHz rUR =0.88 ± 0.04 and for 2800 MHz rUR =0.90 ± 0.03. T hat is the frequency 1415 MHz is optimal. But the so lar activit y excites much more close (to the detector) the process of geomagnet ic act ivit y and it is legit imately to speculate that latter is direct cause of t/variation. The main extremum of correlation is almost at the same time shift (about 10 more
__ ®dV ____________________________________________________________________

for the given case), but it is weaker: rUDst = -0.87 ± 0.04 . Correlation time asymmetry is
®dV ret

also weaker: max | rUDst | / max | rUDst |= 1.11 variations are excited just by solar activit y, relat ion (negat ive by nature of Lto-index) 1415 MHz the main extremum rDstR = -0.

± 0.06 . On the other hand, though the Dstdue to complexit y of their relation, their coris rather weak. For given series Dst and R at 38 ± 0.07 is observed at = -10 (Dst is re-


tarded relat ive to R). Thus we have 7^=0.92 ± 0.03, rUDst = -0.87 ± 0.04 (both advanced) and rDstR = -0.38 ± 0.07 (retarded). Such relationship suggests that connection of U and R is


direct, i.e. nonlocal. But all three links might be nonlinear. Indeed nonlinearit y of (classical local) R ­ Dst link is well known, as well as Dst ­ U (Sec. 9) and R ­ U [11, 14, 16]. But independence functions are equally fit for linear or any nonlinear t ype of dependence and we let Z = R, Y = Dst and X = U. The fulfillment of Ineq.(10) is sufficient condit ion for localit y o f connect ion alo ng t he causal chain RDstU (since any local solar influence o n the detector can not come avo iding the magnetosphere that is source of Dst variat ions). All three independence funct ions of Ineq. (10) were calculated wit h ment ioned above t ime shifts. For estimat ion o f their st abilit y all three serieses were alternately no ised by 21% (by power) flicker-noise [14]. The results are: iU|R = 0.46+-00..0012 , iU|Dst = 0.51+-00..0002 , iDst|R = 0.83+-00..0002 . Ineq. (10) is reliably violated, therefor e con necti on RU is nonlocal. Eve n choice of optimal frequency of R 1415 MHz is not crucia l: for 610 MH z iU|R = 0.50+=00..0031, for 2800 MHz iU|R = 0.49 +-00..0021 , I neq.( 10) is violated, though sli ghtl y le ss.

9

Quantitative Interpretation

Taki ng into account complexity and, as a rule, poor knowledge of large-scale natural s ourcepr ocesses par ameters, it is e xtre mely diffic ult to verify theoreti cally values of time shifts by the detector signal and standard geophysical data. But it is possible to hope on order estimation of s in Eq. (6), i.e. on veri fication of effe ct magn itude. The pr ocess of ge om agne tic ac tivity is the most con ve nie nt, because it admits to use in the right -han d site of Eq. (6) the simplest mode l for the source entropy production densit y: |Z(f)|2 s&= = , (11)

rkqrkq

where E is electric fie ld, f is frequency, r is med ium resist ivit y, q is med ium temperature, Z is impedance, F is magnet ic fie ld. The Z and r we may consider for simplicit y as scalars. By subst ituting Eq. (11) into Eq. (6) further simplification is possible, using the known properties of the electromagnet ic field o f the magnetospheric source. First, the field F is well approximated by plane wave, therefore it is possible to factor out the s& from the integral, and, restricting our consideration to the spectral amplitudes, we reduces this integral simply to thickness of dynamo -layer. Second, use quasi-steady-state approximat ion of the plane wave impedance of homogenous medium: | Z( f ) |2 = 2pfm0r .Dependence on r disappears, and for spectral amplitudes it is easily to show [11, 14, 17, 18] that following ratio is frequency-independent: U( f )
77 2

(/)

/

= const

(12)


and analogously for / (f). Of course, Eqs. t ype o f (12) are approximated, because above simplest expressio n for |Z (j)| is rather rough approximat ion. But the geomagnet ic act ivit y, as a separate source process, has a flaw - it is clo se correlated with solar activit y especially at long per iods T > 27 days. On the other hand, short periods (and correspondingly small space scales) T < 1 day do not cause enough strong detector reaction. It holds significance also choice of an index of geo magnet ic activity. As it had been shown in the previous studies the most effective was to correlate the detector signals not with the magnet ic field measured at the setup site (although it was possible [11, 14, 17, 18]), but with Lto-index of global geomagnetic activit y, which corresponded to t he most large scale electric current system in t he magnetosphere [11, 14 - 16]. Dst-index due to procedure of its calculation is most representative at T > 2 days. F or these reasons the spectral window 20 > 1 > 2 was selected for analysis. However in that window nonlocal int erference fro m the synoptic act ivit y is just possible. Therefore it is a need to select for analysis enough lo ng t ime segment with quiet weather condit ion. That is why in the all previous studies we succeeded in est imation o f <7only in one case [18]. It was estimation by electrode detector and setup's quantum magnetometer data:
10 Forecasting Applications
Availabilit y of essential advanced detector respo nse on natural large-scale dissipative processes gives sense attempt of per formance of the forecast problem. As it is seen in Figs.1, 3, 4 relation of the probe and source processes is far fro m 8-correlated (the do-


main of advanced dependence is spread in wide t range). Therefore forecast algorithm must be based on plural (perhaps, nonlinear) regressio n ­ one forecasted value is calculated as convolution of impulse transition charact erist ic with mult itude of the preceding detector signal values. In addit io n, the problem of optimal filtration, suppressing interference fro m nonlocal influence of other natural pro cesses must be previously so lved. Elaboration of such algorithm is complicated though quite standard task. For the present we will confine ourselves by w ittingly pr imit ive simplest demo nst ration o f the forecast possibility by the example of geomagnetic activity. We select the highest observed correlation pike of optimal filtered time series (Fig 3) and shift t hem on corresponding t.

Figure 5. The detector signal U(mV) forecasts the random co mponent of geomagnet ic activit y Dst (nT) with advancement 42 days. T he or igin o f t ime count corresponds 5/10/1995. In Ref. [11 ­ 14 ,16] t he number o f examples of the so lar (advancement 42 ­ 130. days), synoptic (advancement 73 days) and other geomagnet ic (advancement 33 ­ 130 days) forecasts are presented. The showed in Fig. 5 and cited forecasts are background statistical ones. As for some individual events, our experience had sho wn that detector respo nded only to the most powerful o f them, such as so lar flares of X-class [15]. Visible detector signal is very


smooth usually. But so met imes, e.g. at the beginning o f 2003 extremely sharp splashes (with duration of order of hour) were observed in the detector signal. The biggest splash (134 ± 0.5 µV) was on February 3. And just 42 days after, the famous so lar flare o n March, 17 happened. It was seldo m, gigant ic flare of class X. In such a manner this powerful so lar event caused advanced response of the electrode detector. Moreover splash shapes o f the self-potentials and solar X ­ rays one [15] are similar. In spite o f great est magnitude that so lar flare was not geoactive (it did not cause a global magnet ic storm because o f its inappropriate position on the Sun). Therefore influence of this so lar event o n the detector signal was direct, i.e. nonlocal.

11 Conclusion
The long-term experiments have confirmed existence of nonlocal transaction of so me large-scale dissipat ive processes. The most prominent property of this pheno menon is transaction in reverse t ime. It gives the possibilit y of observat ion o f the future noncontrolled by an observer. This conclusion is consequence of the experimentally verified fact, that nonlocal correlat ion not only vio lates Bell-t ype inequalit y, but also strong causalit y. It has been demonstrated by the successful forecast of random component of geomagnetic activit y. Of course, presented theoretical approach was essent ially heurist ic and t he mode l might be naive approximat ion o f realit y. Therefore development of the theory at crossing of quantum nonlocalit y, act ion-a-distance electrodynamics and causal mechanics is burning. Nevertheless the effect of macroscopic no nlocalit y can be ut ilized for forecasting and ant icipatory purpo ses at present level of knowledge yet. Acknowledgement This work was supported by RFBR (grant 05-05-64032). The author thanks V.O. Serdyuk, J.V. Gorohov and V.A. Machinin for participat ion in the experiment and data processing.

References
[1] Basharov A.M. (2002) Decogerence and Entanglement by Radiation Decay of TwoAtom System: J. Experimental and Theoretical Phys. V. 121, pp.1249-1260. [2] Cramer J.G. (1980) Generalized Absorber Theory and Einstein-Po dolsky-Rosen Paradox: Phys. Rev. D. V. 22 pp. 362-376. [3] Elitzur A.S. and Dolev S. (2003) Is there more to T?: The Nature of Time: Geometry, Physics and Percept ion. Ed ited by R. Buccery, M. Saniga and W.M. St uckey, Kluwer Academic Publishers, pp. 297-306. [4] Ghosh S., Rosenbaum T.F., Aepll G.A. and Co ppersmit h S.N. (2003) Entanglement Quantu m State of Magnetic Dipoles: Nature. V. 425, p.48. [5] Home D. and Majumdar A.S. (1995) Inco mpatibilit y Between Quantum Mechanics and Classical Realism in the Strong Macroscopic Limit: Phys. Rev. A. V 52, pp. 4959-4962.


[6] Hoyle F. and Narlikar J.V. (1995) Cosmo logy and Action-at-a-Distance Electrodynamics: Rev. Mod. Phys. V. 67, pp. 113-156. [7] Julsgaard B., Kozhelkin A. and Po lsik E.S. (2001) Experimental Long Lived Entanglement of Two Macroscopic Objects: Nature. V. 413, pp. 400-403. [8] Korotaev S. M. (1992) On the Possibilit y of Causal Analysis o f the Geophysica l Processes: Geomagnetism and Aeronomy. V. 32, pp. 27-33. [9] Korotaev S.M. (1995) Ro le o f Different Definit io ns of the Entropy in the Causa l Analys is: Geomagnetism and Aerono my. V. 35, pp. 116-125. [10] Korotaev S.M., Hachay O.A. and Shabelyansky S.V. (1993) Causal Analysis o f the Process of Horizontal Informat ional Diffusio n of Electromagnet ic Field in the Ocean: Geomagnetism and Aerono my. V. 33, pp. 128-133. [11] Korotaev S.M., Morozov A.N., Serdyk V.O. and Gorohov J.V. (2003) Experimental Evidence of Nonlocal Transact ion in Reverse Time: Physical Interpretation of Relativit y Theory. Edited by M.C. Duffy, BMSTU Press, pp. 200-212. [12] Korotaev S.M., Morozov A.N., Serdyk V.O., Gorohov J.V. and Machinin V.A. (2005) Experimental Study o f Macroscopic Nonlo calit y o f Large-Scale Geomagnet ic Dissipative Processes: NeuroQuantology. V. 3, pp. 275-294. [13] Korotaev S.M., Morozov A.N., Serdyuk V.O., Gorohov J.V., and Machinin V.A. (2005) Experimental Study o f Advanced Nonlocal Correlat ion of Large Scale Dissipative Processes: Physical Interpret ation o f Relativit y Theory. Edited by P. Rowlands, BMSTU PH, pp. 209 ­ 215. [14] Korotaev S.M., Morozov A.N., Serdyuk V.O. and Sorokin M.O. (2002) Manifestation of Macroscopic Nonlocalit y in So me Natural Dissipative Pro cesses: Russian Phys. J. V. 45 (5), pp. 3-14. [15] Korotaev S.M., Serdyk V.O., Gorohov J.V., Pulinets S.A. and Machinin V.A. (2004) Forecasting Effect of Macroscopic Nonlocalit y: Frontier Perspectives. V. 13 (1), pp. 41-45. [16] Korotaev S.M., Serdyk V.O., Nalivaiko V.I., Novysh A.N., Gaidash S.P., Gorokhov Yu.V., Pulinets S.A. and Kanonidi Kh.D. (2003) Experimental Est imat ion of Macroscopic Nonlocalit y Effect in Solar and Geomagnet ic Activit y: Phys. of Wave Pheno mena. V. 11 (1), pp.46-55. [17] Korotaev S.M., Serdyuk V.O. and Sorokin M.O. (2000) Effect of Macroscopic Nonlocalit y on Geo magnet ic and So lar-Ionospheric Processes: Geo maget ism and Aerono my. V. 40, pp. 323-330. [18] Korotaev S.M., Serdyuk V.O., Sorokin M.O. and Abramo v J.M. (1999) Geophysical Manifestation of Interact ion of the Processes Through the Act ive Properties o f Time: Phys. and Chem. of the Earth. V.24, pp. 735-740. [19] Kozyrev N.A. (1971) On the Possibility of Experimental Investigation of the Properties of Time: Time is Science and Philosophy. Edited by J. Zeman, Academia, p.111132. [20] Kozyrev N.A. (1980) Astronomical Proofs of Reality of 4D Minkowski Geometry: Manifestation of Cosmic Factors on the Earth and Stars. Edited by A. A. Efimo v, VAGO Press, pp. 85-93. [21] Prigogine I. (1979) From Being to Becoming. Published by W.H. Freeman & Co.