An evidence for solar activity influence on the meteorological processes in the south polar region of Mars during the great opposition in AD 1924
AAN EVIDENCE FOR SOLAR ACTIVITY INFLUENCE ON THEMETEOROLOGICAL PROCESSES IN THE SOUTH POLAR REGION OFMARS DURING THE GREAT OPPOSITION IN AD 1924 Boris Komitov
Bulgarian Academy of Sciences – Institute of AstronomyBulgaria, 6003 Stara Zagora-3, POBox [email protected]
A time series of the Martian south ice polar cap mean diameter for the periodJuly-December 1924 is investigated. The data are based on the high quality pictures , whichare obtained by visual observations of 60-cm telescope in Hamburg Observatory during thegreat opposition of Mars in AD 1924. After removing of the seasonal trend (caused by thespringtime regression of the cap) quasi 36 and 80-82 –days cycles in residuals has beenobtained. The sunspot activity spectra for the corresponding period is almost the same one.The local maximums of polar cap area residuals has been occured ~ 10 days after thecorresponding minimums of sunspot activity. The so obtained results are briefly discussed.Keywords: Sun-climate relationship, Mars , polar caps
One of the most interesting and discutabile problems concerning the Earth’sclimate is about the role of solar activity for the changes of the last one. Accordingthe present point of view the climate of our planet depend by large number offactors, which influence is very dificult to separate each from other. By this fact , thestudy for climatic changes of other planets in Solar system , the searching forrelationships with the solar activity may give an impotrant additional contribution tosolving of the above mentioned problem. For this aim Mars is the most suitable object. The planet is observed more than300 years and since the 60’s it is investigated by using of space probe basedinstruments. The processes of building and regression of polar caps , the duststorms , the generating and evolution of clouds and aerosol structures inatmosphere, the seasonal variations of color and albedo in separate regions of thesurface are giiving an important information about the metheorology and climate ofthis planet. Close to this moment the most certain indicator of Martian climatic changes arethe polar caps dynamic data. They are observed from the 17 th century. At thebeginning of 20 th century there are already many pictures on the base of visualobservations, where has been provided by many researchers in differrent times.During the last ~100 years a large number high precission photographic andmicrometric observations has been added. During the last two decades the “weightcenter” is already shifted to CCD-camera observations /ground based , from spaceprobe boards as well as from the Hubble Space Telescope. Since 1997 the the space probe orbiter modules observations are alreadypractically continuous. This is yet not enough for study of processes , where are related to the polarregions climate changes. Although the observations from Martian satelites showsclear evidences, that the Martian climate and especially in near polar regions issubjected of significant changes. For example the layer structure of the summeresidual (water) north polar cap most probably is caused by different regressionrates and , connsequently , by interleaving of warmer and coulder climatic epochs .However the ground based telescopic observations in 20 th century for climatecalled polar caps regression rate changes remain the main important informationsource on this stage.Antoniadi [1] is the first researcher, who conclude that the polar capsregression rates are depended by solar activity level. On the base of visual drawingsanalysis for the period 1852-1912 he established that the polar caps regression ratesare smaller during the years of solar sunspot activity minimums as in the years ofmaximums. Bassu make similar conclusions [2]. He analysed the observation data/mainly photographic inmages/ for the period of 1905-1988. However theobservational material which has been used in the both abovesaided studies is veryunhomogenious. The differrences are caused mainly by the different methods,telsescopes, individual features and experience of observers as well as by thedifferrent conditios of Mars visibility. Some phenomenas such as dust storms ,aerosols or clouds in polar regions can be load to significant errors. *Although theconclusion about the revert dependence between the solar activity level and polarcaps regression rates seems to be logical, because the coulder climate /slowregression/ must corresponding to lower solar radiation levels.Obviously the solution of this problem is on its initial stage. The using oflarger and larger number of observations from orbiting arround Mars space probeswill play the most important role in the near future.
2. The aim of present study. Data and initial proceedings
A typical feature of ground based observations of Mars during the 19 th and 20 th centuries is in the fact , that they are very intensive when the so called “greatoppositions”occurs. With most higher probability during these periods is possible toobtain series of drawings and photographies from one and the same observers, byusing of one and the same telescopes, observational technique and imageproceeding trough short time intervals between separate observations and during offew months periods.It could be used such image series for metheorological conditions trackingover chosen regions of the planet during the period of observations. If (eventually)short-time quasi-cyclic variations will be obtained , a comparison with the short –time variations of solar activity will be reasonable. An other aim in this course can bedetermination of the “resident time” of the solar activity level changes influenceover processes in near-surface atmosphere.Such homogeneous series ,containing 38 qualitative drawing images of Marshas been obtained by Kazimir Graff in 1924 by using of 60-centimeter refractor inHamburg observatory[3] (fig.1). The observations has been provided during theperiod July-December of the same year. The south polar cap is clear shown. Theperiod of observation contain the end of winter and the beginning of spring in thesouth Martian hemisphere. The last one give for Graff the possibility for tracking ofthe mean angular diameter d of south polar cap during the period of its melting .On the base of smoothing data curve the values of D has been determinedby step of 5 days. The last one correspond approximately to the mean time intervalbetween two adjacent observations. The so obtained time series is an object of ournext analysis. Fig1. Drawings of the South polar cap melting in 1924 by K.Graff (published in [3])
The daily smoothed sunspot activity index Ri /the International Wolf’snumber/ for the period 1818-2004 could be downloaded by Internet at address: ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SUNSPOT_NUMBERS . The part ofvalues , which correspond to the Graff’s period of observations has been averagingfor 5 days. Thus the so obtained time series of sunspot activity correspond to themean south polar cap angle diameter series.The both derived final modified time series of Ri and D has been analysedfor existing of trends and cycles. . The T-R periodogramm analysis
In the present work a numerical procedure, called “T-R periodogrammanalysis” (TRPA) for detection of cycles in time series has been used. Comparingwith the “classical “ Fourier-analysis and the large part from other usualy usedmethods, it is essentialy more comfortable and sensitive for detection of cycles whenthe ratio
N/T is non-integer value in the common case. ( N is the time series lengthand T- the period , which corresponded to the length of cycle). For first time TRPAis described in [ 4]. A more detailed and developed version has given in [5]. Itcontain the follwing steps: 1. An approximation of the investigated time series by using of the least squareprocedure over a series of simple periodic functions : f ( t ) = Ao + A cos ( 2 ππ t/T) + B sin ( 2 ππ t/T) ( 1)where Ao is the mean value of all terms in time series and the coefficients A and B can be determinate by solution of the system: (2)where y i =F(t) are the terms of the time series and the corresponding moments are t = 0, 1, 2…. N - 1. The time interval of “1” between the adjacent terms is the “timeseries unit”, i.e. the step of the time series. It real value may be equal of year, day ,minute e.t.c. /In our case this step is equal of 5 days./The period T is varied by equidistant step ∆∆ T from choosed value T to somemaximal one T max . The lower possible limit for T is 2.2. For every one of the so obtained simple periodic minimized functions thecorresponding coefficient of correllation R with the time series, as well as its errorare calculated. The obtained series of R values (T-R –corellogram) have localmaximums near this values of T , where corresponded of potentially existing cycles inthe time series. The amplitude (power) of the cycle may calculated by formula: (3)3. Tests for statistical validity of the cycles. A two criteria are used there.According the first one it need R/SR >1.96 . However very often in pseudo-randomnumber series are generated cycles, where the abovesaid test is covered. Because ofthis fact in [5] a second, more hard test is oferred. It has been established on the baseof analysis of more than 5000 pseudo-random number series. According the last oneit needed
R/SR > 4.54/ N +3.5 for “non-random” generating of cycle . It is clear thatfor enough long series (N -> ∝∝ ) the “critical” value of R tend to 3.5. In the caseswhen the R value is between the both “critical” limits the problem about the cyclevalidity is solved on the base of other expert estimations.
4. Quasi 36-37 and 80-days oscilations of mean temperatures in south polarregion of Mars during the second half of 1924 th As it shown on fig.2. the time sereis of D /circles/ can be separate on threeparts: 1. The most left presented the dynamic of south polar cap dimension in July1924. It characterized by slight decreasing of D , i.e. monotonic melting ofthe polar cap. SR RN == −− ( ) cos ( ) cos ( ) sin ( ) cos ( )( ) sin ( ) cos ( ) sin ( ) sin ( ) y Ao iT A iT B iT iTy Ao iT A iT iT B iT i iNiN iNi iNiN iN −− −− == −− ++ −− −−−− −− == −− −− ++ −− ==== ====== == ∑∑∑∑ ∑∑∑∑∑∑ ∑∑
211 111 1 2 ππ ππ ππ ππππ ππ ππ ππ a T A T B T ( ) ( ) ( ) == ++ . A phase of fast regression /melting/ which started aproximately at thebeginning of August . This process is not light and two “waves” oftemporal increasing of D , i.e. recovering of the ice cap are shown.3. Near to end of November the regression has been practically stoped andthe mean polar cap diameter D remain near to 10 °° .By using of a regressional procedure tests for choosing of the best fit functionfor decribing of the polar cap diameter dynamic as non-linear trend has been made.It turned out, that the best expression of the trend is a full polynom of fourth degree(the flat line on fig 2). It is clear shown that the trend line practicaly coincides withthe time series data in regions “1” and “3”. The quasi-cyclic deviations from thetrend are very clear expressed in phase “2”. Phase “3” corresponds of non-volatileresidual /dust mantle/ formation. This stop the further sublimation of “dry” ice(CO ) as well as the destroying of polar cap too.This behaviour of D is in good agreement with hypotesis , that in thereached Martian surface solar radiation exist a component , which by one hand isstrong absorbed at large solar zenith angles and by other it caused strong quasi-cyclic variations with near-month duration. The solar UV-radiation at λλ≤≤
180 nm issatisfied to the both features. However its absorbtion in the low Martian atmosphereis very strong and it almost not reached to the surface. Consequently ,it remain onlythe possibility that the abovesaid quasi-cyclic solar component is the high energeticcorpuscilar radiation. As is well known the basical near-monthly cycle of theobserved solar activity is by duration of 26-28 days during the decreasing phases of11-year solar cycles. It is almost one and the same by observations both from Earthand Mars . However, during the increasing phases very often the quasi 28-28 dayscycle is weaker relatively to oscilations with duration larger than mont
Fig.2 The mean angular diameter D of Martian south polar cap (by darkcircles) (Jul-Dec, 1924) and the seasonal trend (the flat line) , which is approxymatedby a full polynom of fourth degree. hat is why to be verify the hypothesis of the solar determining of the meandiameter variations for the south polar cap in the phase “2” - an analysis has beenmade on the base of the following steps :1.
Removing of the non – linear trend from the time series of the valuesconcerning D .2. Removing from the new established series the residual variations of thevalues , refer to the phases “1” and “3”.3.
The reduction time series of the residual variations d , related to thephase “2” and containing totally 21 values / for the period of time 110days / has been investigated for the presence of statistical reliable cyclesby means of T-R periodogramm analysis [4].4. The so obtained T-R corellogram for the series of the residual variationsof the polar cap mean diameter has been comparing with thecorresponding one , but derived for the time series of the avarage for fivedays values of the sunspot index Ri for the same time .The results are shown on Figs. 3 and 4 . It is evidently that at the first of them in thevariations of the mean diameter of the south polar cap , during the period August –November 1924, presents a good showed 36 –day cycle as well as and a second one,weaker – but reliable quasi 80 - days cycle .The last one is resonancely multiple ofthe “classical” 27 – solar cycle .It is clear shown from fig.4 that the main near-montlhly oscilation of solaractivity level is by duration of 36-37 days. It is almost equal to the basic cycle of thepolar cap variations. The traces of 27-days cycle are weak, but in addition there isvery clear visible and statistical valid quasi –80-days oscilation in sunspot activity.Consequently, the both correllograms in fig.3 and 4 are very similar. This indicatethat a strong relationship /may be anti-corellation/ between d and Ri may to expect.The searching for direct relationship between these both parameters pointout, that the maximal by module correllation coefficient r = - 0.73 occur when thephase shifting is 10 days . Ri overtackes d , i.e. the maximal polar cap rebuildingoccur about 10 days after the local minimum of the sunspots. The statistical validityof this relationship exceed 99.9%.The physical explanation of this phenomena could be search in two courses:1. The penetrating to the surface solar variable component /protons by energy>1 MeV/ reach local maximums near to the maximal levels of sunspot activity.The last one lied to corresponding changes of near surface atmosphereincreasing the aerosol production rate and consequently- to temperature andpolar cap regression rate variations. The “resident time “ of this process is about10 days. The observed cases of particular polar cap rebuilding are correspondingto maximal cooling of atmosphere. Then the CO “dry ice” sublimation rate isgoing down and the condensation is more active.2. During the near sunspot minimum conditions as in AD 1924 an othermechanism could be much more probable. It is related to the increasing of thefalling in Martian atmosphere galactic cosmic rays (GCR) flux. The highervalues of the last ones corresponds to low level of the solar wind flux parameters(the Forbush -effect). Thus if any solar active center is going out of view fromthe Mars possition it should be follow an decreasing of solar wind and increasingof GCR flux near to the planet 3-4 days later. As in the case “1” the GCR fluxncreasing should to intensifed the aerosol production rates . This mechanismhas been firstly described by Svensmark and Friis- Cristiensen [6]. It should alsotaken into account also a possible “resident time” which need to generate ameteorological effect of cooling in the atmosphere.It should be note , that there is also an uncertainity, which come from thedifferent observational conditions of the Sun from the Earth and Mars. While thepublished Ri data are Earth- related , i.e the observational conditions from the Marsposition in one and the same moments should be different. It concern the solar windand GCR –flux conditions in the Martian near space too. However the possiblediffernce by the above mentioned geometrical causes should be not very large,because of the fact of the Mars opossition in the second half of 1924 th . On otherhand it is absolutely possible that a significant part of the 10 days shifting in therelationship between Ri and d has been caused namely by this geometrical factor.Thus by the observation geometry as well as by the physical causes which aredescribed in “1” and “2” it is very difficult to discussed of the relationship shiftingin some certain physical aspect. It is only an observational phenomena in this study. Fig 3. The T-R corellogram of the d –residuals (August 1 –November 20, 1924)ig 4. The T-R corellogram of the dayly Inyernational Sunspot Number (Ri)series (August 1 –November 20, 1924)
5. ConclusionThe obtained results shown that homogeneous photographic images series,obtained during the “great oppositions” in 20 th century in relatively short timeintervals / few days/ can be used for study of short periodic metheorologicalparameters oscilations in choosen regions of Mars, such as the polar caps. It isevidently of the all things that, the role of acive processes on the Sun over themeteorology and climate of this planet is significant.It can be continued the investigation by using of additional photographicmatherial from ground –based observations, but first of all- on the base of spaceprobes information.REFERENCES1. Antoniadi, E. M.
Basu D.,
Kovachev B,
Komitov B.,
Komitov B.,