Wyniki wyszukiwania - BazTech

1

Związki początku termicznego okresu wegetacyjnego w Polsce z anomaliami temperatury powierzchni wody na Atlantyku Północnym (problem prognozy długoterminowej)

Marsz A. A., Żmudzka E.

Przegląd Geofizyczny

|

2002

|

Z. 3-4

165-201

EN

The paper presents the results of the inquiry into the relations between the dates of the beginning of the meteorological growing season in Poland and the thermal state of the surface of the Northern Atlantic. The variability of this state (the magnitude of the streams of heat flowing from the ocean to the atmosphere, and the distribution of the areas in which the heat transmission is greater or smaller than on the average) conditions the variability of the atmospheric circulation within the reach of influence of the Northern Atlantic, which, in turn, shapes the variability of the beginning of the growing season in Poland. The properties of the distribution of the anomalies of water temperature in the Northern Atlantic were characterised with 10 representative fields, the ,,grids", while the beginning of the growing season was characterised with the sequences of dates (being the numbers of the day in a year corresponding to the instance of passage of the average daily air temperature through the value of 5°C) from 34 weather stations located in Poland (except for the mountainous areas) in the years 1970-1998. The basic method applied in the study is the correlation analysis. The quasi-synchronous relations were analysed (the dates of the beginning of the growing season were correlated with the monthly anomalies of the SST in the particular grids for the same year), as well as the synchronous ones (the shift of the series analysed by one year: the SST from the feth year, the beginning of the growing season - from the k + 1st year). Dependence of the date of the beginning of the growing season on the SST anomalies in some reference grids was established. The dependencies identified are in their majority the asynchronous correlations, and the spatial distribution of the isocorrelates is specific for a definite grid. It was also stated that a more important role in the shaping of the beginning of the growing season is played not so much by the thermal conditions within the particular water areas as by the meridional water temperature gradients between these areas either in a given month or on the average for a period of several months. Attention should be paid to the conformity of the signs of the correlation coefficients for the beginning of the growing season and the SST anomalies appearing a year before, observed synchronously and preceding by 2-3 months the starting instance of the season, and to the distinct spatial and temporal order in the distribution of the SST anomalies, which influence strongly the dates of the beginning of the growing season. The most frequent are the associations with the thermal conditions in the water areas of the Western part of the Northern Atlantic (the Sargasso Sea and the areas to the north of the Gulf Stream delta), with the precedence of the date of the beginning of the growing season amounting to between 11 and 14 months. As time passes, the influence of these water areas fades away and the correlations appear indicating a smaller time precedence (4 to 9 months), concerning the ocean areas located in the moderate latitudes of the central and Eastern parts of the Atlantic Ocean. The article offers, as well, an attempt of a physical interpretation of the results obtained. The paper presents the results of the study aiming at elaboration of the method allowing for the long-term forecasting of the beginning of the growing season for the lowland and upland Poland on the basis of water temperature anomalies of the surface of the Northern Atlantic (their values and spatial distribution). The analysis was carried out with respect to the series shifted by one year: the STT anomalies for the year k, and the beginning of the growing season for the year k + 1. Th study of the nature of relations made it possible to establish that the interdependencies appearing can be described sufficiently precisely with a multivariate linear regression model. In the regression analysis, which was limited - in view of the stability of the regression coefficients - to three independent variables, attention was paid to the relations, which can be physically explained and to those, which simultaneously explain more than 50% of variance of the beginning of the growing season over a broader area, and at least 60% in the zone of the strongest influence. The respective equation has, as well, to satisfy in the zone of appearance of the strongest influence the criterion ofp < 0.001, and the statistical significance of all the three estimates of the directional coefficients of the equation atp < 0.05. Resulting from the multivariate analysis conducted were two prediction equations, fulfilling the prerequisites set. Each of the two equations is proper for a definite part of Poland: 1. Western and Central Poland, 2. Northern and Eastern Poland. These equations contain both the raw data (not normalised monthly averages of the values of the SST anomalies) and the combined variables (like the averages of the values of SST anomalies for the definite water areas from the time periods longer than one month, and the differences of the anomalies between the definite water areas, describing the magnitude of the meridional water temperature gradients, as well as differences of the anomalies for the same water areas between appropriate time intervals). Both equations point out that the primary role from the point of view of the influence exerted by individual ocean areas on the dates of the beginning of the growing season in Poland is played by the heat resources of the Sargasso Sea, which control to a large extent the variations of the thermal conditions over the remaining ocean areas. A simplified verification of the prediction equations was also carried out, having shown that their parameters obtained from the 27-year observation series are stable. In view of the fact that the latest SST anomalies entering the equations date from September, the forecast for the beginning of the growing season in the subsequent year can be formulated already in October.

2

Temperatura powierzchni wody na Północnym Atlantyku a temperatura powietrza na Spitsbergenie i Jan Mayen

Marsz A. A.

Problemy Klimatologii Polarnej

|

1999

|

nr 9

37--80

EN

This paper deals with correlation between air temperature over the Greenland Sea and water temperature in the North Atlantic The Jan Mayen and Svalbard-Lufthavn stations have been chosen to illustrate air temperature characteristics for this sea area, The monthly mean air temperature indicates strongest correlations between the above mentioned stations during the polar night (see tab. 1). This fact proves that the element which is common to both stations is atmospheric circulation. Relations between the course of air temperature at these stations and the NAO index turned to be weak (see tab. 2) and their distribution varies widely. NAO has been found not to be the most important element having influence on the formation of air temperature. The analysis of relation between the variability of the sea water temperature in the Greenland Sea and the variability of air temperature at both stations was fruitful as the relations were both weak and not clear enough. The monthly mean water temperature anomalies (SST) in the North Atlantic in grids 2° x 2° over the period January 1970 and August 1997 (26 years and 8 month) were used to analyse correlations between the sea surface temperature and air temperature at both stations. The data concerning the sea surface temperature were the subject of three-stage statistical analysis resulting in 10 grids being distinguished (see fig. 1), each of which is characterised by changes in temperatures of far greater ocean surface. These grids are called 'control grids'. The following symbols are used: ANmm[DD,SS]; where AN - explanatory symbol, meaning that it clenotes anomalies in water temperature in a grid; mm - number of a month when anomalies occur; [DD, SS] - location of a central point of a grid in space, where DO - west longitude, SS .- north latitude. For example symbol AN09[30,54] means that we are dealing with an anomaly SST from September from a grid located 30°W and 54°N. The study of synchronic and asynchronic correlations between SST anomalies in given grids and air temperatures indicated the occurrence of a number of correlations significant from the statistics point of view. They were mainly asynchronic correlations with SST changes proceeded by changes in air temperature. The maximum of air temperature correlations, which are statistically significant, was noted between SST anomalies at the end of winter and beginning of spring (the final winter coding season) and the end of summer and beginning of autumn (the final season of summer warming). It represents two extreme stages from which a further evolution of the area of the water temperature takes place in warm and cold seasons. Figures 2 – 9 show exemplary distribution of correlations coefficient values. As the variability of mea n monthly temperatures in winter months plays the most important role in the dispersion of annual mea n temperature over this area and the less important one in the warmest summer months, the subjects of this study were relations between signs, values and location of SST anomalies and monthly mea n temperatures in winter months (December to March) and summer ones (July, August). A multiple regression was taken as a model and the monthly mean air temperature is a dependent variable and the values of SST anomalies are independent variables. The analysis was carried out with the help of a forward stepwise method. In order to keep the stability of equations the relations only with three independent variable were taken into considerations. Correlations of great statistical significance were found. They allow to estimate monthly mean air temperatures at Jan Mayen and Spitsbergen with the help of SST anomalies which occurred in the North Atlantic earlier. The list of correlation for winter months and their statistical characteristics are shown as equations [1] - [8] and tor summer months as equations [9] _ [12]. They explain 44% to 66% of the variability of temperature in the winter months and from 43% to 60% of the variability of temperature in the summer months. A similar analysis was carried out in order to find relations between synchronic and asynchronic annual SST anomalies in control grids and annual mean temperatures at both stations. These correlations are represented by [13] and [14] equations (synchronic correlations) and by [15] and [16] equations (asynchronic correlations) - the SST anomalies taken from a year ‘n’ the annual temperatures from a year n+1. The variability of annual mean SST anomalies in the chosen control grids explain about 50% ot variabilities of annual mean air temperature in case of synchronic correlations and 22% to 40% in case of asynchronic ones. Figures 10, 11 and 14 illustrate exemplary characteristics of such correlations quality. The analysis of spatial distribution of grids determining the variability of air temperature in the Greenland Sea shows that the temperature over the east part of that sea area (the region ot Spitsbergen) describes the thermal states in the west part of the North Atlantic, mainly the SST anomalies in the Labrador Current, in the Sargasso Sea, in the Gulf Stream and within the cyclonic circulation of the North Atlantic. The air temperature in the western part of the Greenland Sea (the region of Jan Mayen) is also influenced by the distribution of anomalies in the eastern part of the North Atlantic. The formation of the SST anomalies in the Labrador Current in its extreme end see fig. 12) and in the Gulf Stream (see fig 13) has the greatest influence on the course of temperatures occurring over the entire Greenland Sea. The analysis of equations [1] - [16] and table 3 gives more detailed information about the parts of the ocean having the strongest influence on the temperature over the Greenland Sea. The mechanism of these correlations can be explained by modification of atmospheric circulation which is influenced by the flow of warm air from the Nort Atlantic into the atmosphere. The range and spatial distribution of the warm air flowing from the ocean to the atmosphere depend generally speaking, on the heat resources in the ocean, which can clearly be observed in anomalies in surface water temperatures. The increased or decreased mea n streams of heat from a specified part of the North Atlantic when compared to many years ones regulate the value of horizontal thermal gradients in the middle troposphere of medium latitude zones. By stabilising or destabilising long waves (the range in wave number) they have their input into the occurrence of statistically significant regression of location of upper ridges axes or upper trough with all the resulting consequences for atmospheric circulation in the layer of 1000-750 hPa (lower troposphere). Finally, the character of atmospheric circulation in a given region changes (predominance of zonal circulation or one of the meridional type) and this can be foreseeable. Further the changes in atmospheric circulation determine the deviation of climate elements tram the many years' mean values in a given many years' period, year, season or month. The results of this study show that the changes in the character of atmospheric circulation over the Greenland Sea and at the same time in the thermal regime over this sea area have been influenced to a great extent by changes in SST in the region of the North Atlantic in most part of its tropical and sub-tropical zones. The above conclusion makes the thesis that only Arctic generates climatic oscillations which then move to lower latitudes disputable. The results indicate that looking for reasons for changes in climatic conditions in the region of Atlantic Arctica cannot be carried out without taking into consideration the situation in the North Atlantic, especially meridional flow of heat in the oceanic circulation.

3

Związki między temperaturą wody w energoaktywnej strefie Morza Bellingshausena a temperaturą powietrza na Stacji Arctowskiego

Styszyńska A.

Problemy Klimatologii Polarnej

|

1998

|

nr 8

25--46

EN

The main task of this paper is to explain if there is an energy-active sea zone in the vicinity of the South Shetland Islands and the Antarctic Peninsula which controls changes in atmospheric circulation in this area. The analysis made by use of the data comprising information about mean monthly sea surface temperatures (later SST) and SST anomalies in 2 x 2° grids - GEDEX and data about mean monthly air temperatures taken at the Arctowski Station (Meteorological Yearbooks of the Arctowski Station). Common data spanned the period from January 1982 to April 1992. The first stage of this work was to find so called .active grids", i.e. grids of bigger influence of ocean surface on thermic regime of distant areas. In order to do that an analysis of changes in SST in parts of the South Ocean comprising the Bellingshausen Sea, the Drake Strait, the Scotia Sea and the boundary between the Scotia Sea and the Weddell Sea was carried out. The analysis resulted in a conclusion that three grids situated 80oW: 56°,60° and 64°S show the larger relation with the flow of air temperature at the Arctowski Station. There are synchronic and asynchronic correlations between SST anomalies and the air temperature in nominated grids of the Arctowski Station. The results of analysis of synchronic correlations have been presented in table l. Asynchronic correlations are of complicated nature and distributions. Most numerous simple correlations were reported to occur between the temperature at the Arctowski Station and SST Anomalies in grids [80°W, 64°S]. The largest correlations are those with anomalies occurring in January, February and March. They can be observed in the air temperature with 11-13 months delay. The combined correlations are multiple correlations between regression equation of synchronically occurring anomalies (AN) in those grids and the air temperature at the Arctowski Station (ARC) in consecutive months (1, 2, 3, ..., n, n + 1, n + 2); ARC_n = a + b AN[80.56]_n + c AN[80.60]_n + d AN[80.64]_n. Table 2 contains set of multiple correlation coefficients and those which are likely to be significant have been marked. It has been stated that SST anomalies at 800W in March correlate with monthly air temperatures at the end of summer the following year (February and March) at the Arctowski Station and with temperatures of the early and midwinter of the following year (May, June, July).The variation in SST anomalies in March explains 88% - 69% of variance of variation in the air temperature in June and in July of the following year at the Arctowski Station (fig. l). The response of the air temperature to the occurrence of SST anomalies in October at 800W is much faster - from one to five months. Large correlation between the air temperatures at the Arctowski Station and SST anomalies can be observed already in December of the same year and in January, March and April in the following year (fig. 2). The above stated facts lead to conclusion that the distribution of SST does not influence the flow of the air temperature in a continuous way. Future variations in the air temperature are influenced by the states of thermal field of water measured at crucial moments (the end of summer and the end of winter). They are the states, which later on are slowly modified by processes of radiation in-and off flow, wind chilling and dynamic processes active in the ocean (heat advection following the mass advection). Thus a thesis can be stated that the SST anomalies occurring in grids 56°, 600 and 64°S. 800W may serve as predictive values to work out long term prognosis of the air temperature at the Arctowski Station. These prognosis can be divided into "early" prognosis with 2-6 months' advance (equations 1-4) and "distant" prognosis with 11-18 months' advance (equations 5-8). The above mentioned equations explain about 91% to 52% of variations in the mean monthly air temperature at the Arctowski Station. The presented facts indicate that there really is energy-active zone in the Bellingshausen Sea. Chapter 6 in 4 points shows how the hypothetical mechanism works. It can be understood and explained in a similar way as in case of the Labrador Sea and the New Foundland region (Marsz 1997). The analysis of synchronic statistical correlations between the air temperature at the Arctowski Station and the distribution of SST anomalies at 80°W indicates, among others, the presence of the mechanism described in Chapter 6. Such correlations have been analysed and discussed in a detailed way for April (fig. 3, equations 9 and l0) and for July (fig. 4, equation 11).