%K HĹmĂŠrsĂŠklet %P 59-67 %D 2007 %L acta16901 %J Acta climatologica %A Makra LĂĄszlĂł %A KĹvĂĄgĂł TamĂĄs %A Olivie Dirk %O Bibliogr.: 67. p. ; ĂśsszefoglalĂĄs angol ĂŠs magyar nyelven %C Szeged %X ĂsszefoglalĂĄs - A tanulmĂĄny cĂŠlja a globĂĄlis felszĂnhĹmĂŠrsĂŠklet idĹsor nĂŠhĂĄny statisztikai jellemzĹjĂŠnek tanulmĂĄ- nyozĂĄsa. A lineĂĄrsi trendanalĂzis nem mutatott ki hosszĂş tartamĂş szignifikĂĄns trendeket. Az esetleges rĂśvid tartamĂş (legfeljebb hatelemĹą) trendek a legnagyobb gyakorisĂĄggal a D(QTH, QTM) globĂĄlis felszĂnhĹmĂŠrsĂŠklet differencia idĹsor kĂśzĂŠpsĹ harmadĂĄban fordulnak elĹ. A vizsgĂĄlt idĹsorra alkalmazzuk a Makra-tesztet, mely a kĂŠtmintĂĄs prĂłba egy Ăşj interpretĂĄciĂłja. A prĂłba alapkĂŠrdĂŠse: kimutathatĂł-e szignifikĂĄns kĂźlĂśnbsĂŠg a vizsgĂĄlt idĹ- sor egy tetszĹleges rĂŠszmintĂĄjĂĄnak, illetve a teljes idĹsornak az ĂĄtlagai kĂśzĂśtt. A prĂłba alkalmas arra, hogy szignifikĂĄns tĂśrĂŠseket mutassunk ki az idĹsorban, meghatĂĄrozva annak erĹssĂŠgĂŠt, hosszĂĄt ĂŠs tartamĂĄt, azaz kezdĹ ĂŠs zĂĄrĂł idĹpontjĂĄt. A Makra-prĂłba segĂtsĂŠgĂŠvel meghatĂĄroztuk a D(QTH, QTM) globĂĄlis felszĂnhĹmĂŠrsĂŠklet differencia idĹsor szignifikĂĄns rĂŠszperiĂłdusait. KĂŠt szignifikĂĄns rĂŠszperiĂłdust mutattunk ki: PozitĂv tĂśrĂŠst 1861-1900 kĂśzĂśtt, mĂg negatĂv tĂśrĂŠst 1881-1958 kĂśzĂśtt tapasztaltunk. TovĂĄbbi cĂŠlunk az volt, hogy tanulmĂĄnyozzuk a megnĂśvekedett COj szint hatĂĄsĂĄt a globĂĄlis felszĂnhĹmĂŠrsĂŠklet vĂĄltozĂĄsaira. A D(QTH, QTM) idĹsor legjobb gĂśrbeillesztĂŠsei az adatsor varianciĂĄjĂĄnak (teljes nĂŠgyzetes hiba = total squared error) jelentĹs rĂŠszĂŠt megmagyarĂĄrzzĂĄk: R2 > 80 %; RMSE (becsĂźlt szĂłrĂĄs) < 0.5. Az RMSE ĂŠrtĂŠke igen hasonlĂł a referencia szimulĂĄciĂłra kapott RMSE ĂŠrtĂŠkhez. Az adatsorhoz illesztett gĂśrbĂŠk futĂĄsa az idĹszak elejĂŠn nĂśvekvĹ, majd 10-15 ĂŠves futĂĄst kĂśvetĹen elĂŠri maximumĂĄt, s kĂŠsĹbb fokozatosan csĂśkken. Ugyanakkor az illesztett gĂśrbĂŠk szĂĄmos esetben emelkedĹ tendenciĂĄt jeleznek a vizsgĂĄlt 100 ĂŠves periĂłdus vĂŠgĂŠn, mely meglehetĹsen leronthatja a hosszĂştĂĄvĂş elĹrejelzĂŠseket. Summary - The aim of the study was to analyse some statistical characteristics of the global surface temperature time series. Linear trend analysis did not reveal any significant long-range trends. Significant sporadic short term trends (with not more than 6 elements) can only be observed with higher frequency in the middle third of the D(QTH, QTM) temperature difference times. The Makra-test, as a new interpretation of the two-sample test, was also applied to the global surface temperature time series. The basic question of this test is whether or not a significant difference can be found between the mean of an arbitrary sub-sample of a given time series and that of the whole series. This test is suitable to detect breaks in the data set, along with their strength, length and time interval, namely their starting and ending date. With the help of the Makra-test two significantly different subperiods in the global temperature difference time series of D(QTH, QTM) were detected. In the here-mentioned data set a significant positive break were found between 1861-1900 and a significant negative break relating to the term around 1881-1958. A further aim was to study the effect of the increased CO2 level 01. the changes of the global temperature. The curves best fitting to the D(QTH, QTM) global surface temperature difference time series explain changes (total squared error) of the data set in fairly high ratios. The value of the R-square generally exceeds 80 %. The estimated standard deviations (RMSE) are rather low, mostly below 0.5. These RMSE values are very similar to those of standard deviation in the reference simulation. The fitted curves adjusted to the data show an increasing tendency in the first period; then, after 10-15 years run reaches its maximum and later is gradually decreasing. However, the fitted curves in several cases indicate slight increasing tendencies at the end of the 100-year period, which may rather ruin long-range forecasts. %T Global surface temperature time series characteristics for the earth, in relation to CO2 perturbations %V 40-41