Trendi podnebnih spremenljivk in kazalcev

Razlaga kart časovnih trendov podnebnih spremenljivk in kazalcev

Metodologija

Izračunani trendi osnovnih podnebnih spremenljivk (sezonska ali letna povprečja in vsote) temeljijo na preverjenih in homogeniziranih časovnih nizih meteoroloških meritev na opazovalnih postajah. V celoten postopek obdelave podatkov smo zajeli le postaje z zadovoljivo kakovostjo meritev. Več o kontroli in homogenizaciji najdete na spletni strani projekta Podnebna spremenljivost Slovenije: http://meteo.arso.gov.si/met/sl/climate/pss-project/.

Prikazane sezonske in letne vrednosti trendov so aritmetične sredine ustreznih mesečnih vrednosti; za pomlad od marca do maja, za poletje od junija do avgusta, za jesen od septembra do novembra, za zimo od decembra do februarja. Pri snežnih podatkih se letne vrednosti trendov nanašajo na t. i. snežno sezono, ki traja od avgusta do julija.

V izračunu trendov podnebnih kazalcev (število značilnih dni, ekstremne vrednosti) smo upoštevali tiste časovne nize, pri katerih v postopku homogenizacije nismo ugotovili večjih skokov ali trendov, ki bi bili posledica neklimatoloških dejavnikov, hkrati pa so bili dnevni podatki dovolj kakovostni. V nasprotju z metodami homogenizacije časovnih nizov mesečnih podatkov, ki so že izdelane in uveljavljene, so metode homogenizacije dnevnih ali urnih podatkov še v začetnih fazah razvoja in dokaj nezanesljive, zato smo podnebne kazalce namesto na homogeniziranih dnevnih vrednostih izračunali z uporabo izvornih podatkov. Izjema so le posamezni primeri, kjer smo spojili nize različnih postaj (Bilje in Nova Gorica ter Letališče Portorož, Portorož Beli križ in Koper). V teh primerih smo z mesečnim popravkom popravili vse dnevne vrednosti istega meseca.

Linearni časovni trend je izračunan po metodi Theil-Sen. Metoda je znana po zanesljivih rezultatih za asimetrične in heteroskedastičnostne ostanke linearne regresije, a je hkrati le za odtenek manj zanesljiva v primerih, ko so izpolnjene vse predpostavke za uporabo metode najmanjših kvadratov (kar je pri obravnavanih podatkih redko). Več o metodi Theil-Sen najdete na spletni strani.

Statistična značilnost trenda je bila izračunana iz 95-odstotnega intervala zaupanja za trend. Interval zaupanja je bil izračunan po metodi, ki jo je predlagal Sen in je priporočena metoda za izračunavanje intervalov zaupanja trendov po metodi Theil-Sen (Sen, P. K., Estimates of the regression coefficient based on Kendall's tau, Journal of the American Statistical Association 63: 1379–1389 (1968)). Trend se razlikuje od 0 pri stopnji zaupanja 5 %, če ima ves interval zaupanja isti predznak. Mali krogci predstavljajo statistično neznačilen in veliki krogci statistično značilen trend pri 5 % stopnji zaupanja.

Opisi podnebnih spremenljivk in izpeljanih kazalcev

Temperatura zraka

Temperaturo zraka merimo v meteorološki hišici na višini 2 m, s termometrom na desetinko stopinje Celzija natančno. Dnevno povprečno temperaturo izračunamo iz treh terminskih meritev ob 7., 14. in 21. uri po sončnem času, in sicer po naslednji enačbi:

T= (T7 + T14 + 2 ∙ T21) / 4

Mesečna povprečna temperatura je vsota vseh dnevnih povprečnih temperatur deljena s številom dni v mesecu.

Dnevno najvišjo in najnižjo temperaturo merimo v meteorološki hišici na višini 2 m s posebnim maksimalnim oziroma minimalnim termometrom; temperaturo odčitamo enkrat dnevno, ob 21. uri po sončnem času. Mesečno povprečje izračunamo z deljenjem vsote dnevnih vrednosti s številom dni v mesecu.

Značilni dnevi so določeni z najvišjo ali najnižjo temperaturo zraka. V zelo vročem dnevu najvišja temperatura preseže 35 °C, v vročem 30 °C in v toplem 25 °C. V ledenem dnevu je najvišja temperatura pod lediščem. Hladen dan je določen z najnižjo temperaturo pod 0 °C in mrzel dan z najnižjo temperaturo pod –10 °C. Tropsko noč beležimo, ko je dnevna najnižja temperatura nad 20 °C.

Padavine

Količino padavin merimo s Hellmannovim dežemerom (pluviometrom) ob 7. uri zjutraj po zimskem času. Dnevna višina padavin je vsota padavin, ki so padle od 7. ure prejšnjega dne do 7. ure dneva meritve, pripisana je dnevu meritve. Mesečna višina padavin je vsota vseh dnevnih vrednosti v mesecu. V primeru, da gre za padavine v trdnem stanju (sneg, toča, … ), jih izmerimo tako, da padavine pri sobni temperaturi počasi stalimo in izmerimo nastalo vodo.

Največja sezonska ali letna vrednost dnevne ali dvodnevne višine padavin je največja od dnevnih ali dvodnevnih vrednosti v določeni sezoni ali v letu.

Število dni s padavinami vsaj 20 mm (50 mm, 100 mm), dobimo tako, da preštejemo dneve, ko je padlo vsaj 20 mm (50 mm, 100 mm) padavin.

Novi in skupni sneg

Višino skupne snežne odeje merimo ob 7. uri zjutraj po zimskem času na travnati površini. Višina nove oziroma sveže snežne odeje, pomeni višino snega, ki je zapadel od 7. ure prejšnjega dne do 7. ure dneva meritve, in je pripisana dnevu meritve.

Mesečna povprečna višina skupne snežne odeje je vsota dnevnih vrednosti deljena s številom dni v mesecu. Mesečna vsota novega snega je vsota dnevnih vrednosti.

Največja sezonska višina skupne snežne odeje je največja od dnevnih vrednosti.

Sončno obsevanje

Trajanje sončnega obsevanja merimo s heliografom, ki je sestavljen iz krogelne leče in podstavka, na katerem je pritrjen registrirni trak, heliogram. Krogelna leča zbira sončne žarke v svojem gorišču in izžiga sled na heliogramu. Z obdelavo heliogramov določimo urne vrednosti trajanja sončnega obsevanja in s seštevanjem urnih vrednosti dobimo dnevne, mesečne in letne vrednosti trajanja.

Potencialna evapotranspiracija

Potencialna evapotranspiracija ni merjena spremenljivka. Izračunamo jo iz drugih meteoroloških spremenljivk. Po priporočilih FAO dnevno vrednost potencialne evapotranspiracije izračunamo po Penman-Monteithovi formuli za referenčno površino trave. Pri tem upoštevamo odvisnost izhlapevanja od energije globalnega sončnega obsevanja, temperature in vlažnosti zraka ter hitrosti vetra. Vse vhodne spremenljivke smo pred izračunom dnevne evapotranspiracije preverili in časovne vrste podatkov o vetru prilagodili na zadnje obdobje, ko je bila hitrost vetra merjena na samodejnih meteoroloških postajah. Mesečna vrednost potencialne evapotranspiracije je vsota vseh dnevnih vrednosti v mesecu.

Zračni tlak

Zračni tlak na opazovalnih postajah smo v preteklosti večinoma merili z živosrebrnim barometrom, danes pa ga merimo z elektronsko napravo. Pri meritvi z živosrebrnim barometrom opazovalec odčita temperaturo na barometru in višino živosrebrnega stolpca. Iz obeh podatkov izračuna zračni tlak na postaji. Dnevno povprečje tlaka je aritmetična sredina vrednosti ob 7., 14. in 21. uri po zimskem času.

Comments to the maps of trends of climate variables and indices

Methodology

The trends of basic climate variables (seasonal and annual means and sums) are based on quality controlled and homogenised time series of meteorological measurements at observing stations. Only the stations with an adequate quality have been subjected to the whole process of data analysis.

Depicted seasonal and annual trend values are the arithmetic mean of corresponding monthly values. Spring spans March through May, summer June through August, autumn September through November and winter December through February. Annual trend values of snow data refers to snow season period, from August till the end of July in the following year.

The selection of time series for the calculation of climate indices (number of days above/below threshold, extreme values) is based on two criteria: no large breaks or trends in the time series due to non-climatic effects found during homogenisation process and adequate quality of daily data. While the homogenisation methods are well developed and recognized for monthly data, the methods for daily and hourly data are still in its infant phase of development and thus unreliable. Therefore the climate indices have been calculated on original rather than homogenised daily data. The only exceptions to the rule are merged time series from two or three different stations (Bilje and Nova Gorica as well as Portorož Airport, Portorož Beli križ and Koper). Monthly corrections have been applied to the daily values in these cases.

Linear trend in time series is calculated by Theil-Sen method, which is known for its robustness for asymemtric and heteroscedastic residuals in linear regression. On the other hand it is hardly less reliable in the cases where all conditions for the least-square-method are met (which are seldom in our case). More about the Theil-Sen method may be found on Wikipedia .

Statistical significance of the trend is based on the 95 % confidence interval. The interval was calculated by a method proposed by Sen and is also recommended for computing the confidence interval of the Theil-Sen estimator (Sen, P. K., Estimates of the regression coefficient based on Kendall's tau, Journal of the American Statistical Association 63: 1379–1389 (1968)). The trend is different from zero at 5 % significance level if the sign of the whole confidence interval is the same. The small circles denote statistically insignificant and the large circles statistically significant trend at the 5 % level, respectively.

Description of climate variables and derived indices

Air temperature

Air temperature is measured inside Stevenson screen, 2 metres above ground, using a thermometer at resolution of a tenth of a Centigrade. Daily mean value is calculated from three measurements, at 7 a. m., 2 p. m. and 9 p. m. local time according to the following formula:

T= (T7 + T14 + 2 ∙ T21) / 4

Monthly mean temperature is the sum of daily means divided by the number of days in given month.

Daily maximum and minimum temperature is measured inside Stevenson screen, 2 metres above ground, with a special maximum and minimum thermometer, respectively. Both temperature extremes are read once per day, at 9 p. m. local time. The monthly mean is calculated by dividing the sum of daily values with a number of days in a given month.

Climate indices of daily air temperature extremes are defined by maximum or minimum air temperature. A very hot day corresponds to a maximum temperature of more than 35 °C, a hot day to the maximum of more then 30 °C and a warm day to the maximum of more then 25 °C. Icing day is when the temperature stays below 0 °C for the whole day. Frost day is defined by a minimum temperature below 0 °C and a very cold day by the minimum below –10 °C. Tropical night is recorded when the minimum temperature exceeds 20 °C.

Precipitation

Precipitation amount is measured daily at 7 a. m. CET by Hellmann rain-gauge (pluviometer). The daily precipitation sum corresponds to the interval from 7 a. m. of the day before to 7 a. m. of the measurement day and is assigned to the measurement day. Monthly value is the sum of daily values in a given month. Solid precipitation (snow, hail …) is melted slowly at room temperature before the amount of water is measured.

The maximum seasonal and annual values of daily or two-days precipitation sum corresponds to the highest of all values in a given season or year, respectively.

By counting the days with a daily precipitation sum of at least 20 mm (or 50 mm or 100 mm) we got the seasonal or annual values of such days.

New and total snow

The total snow depth is measured on a grass field at 7 a. m. CET. The depth of new or fresh snow cover corresponds to the snow, which has fallen from 7 a. m. of the day before to 7 a. m. of the measurement day and is assigned to the measurement day.

Monthly mean of total snow depth is the sum of daily values divided by the number of days in given month. Monthly sum of new snow is the sum of the daily values in a given month.

The maximum seasonal total snow depth represents the highest of the daily values.

Sunshine

Bright sunshine duration is measured by a sunshine-recorder, constructed from spherical lens and a mount, where card, heliogram, is fixed. The lens focuses solar rays in the focal point on a card, creating trace. Hourly sunshine duration values are determined by an analysis of heliograms. By summing up the hourly values we got values of daily, monthly and annual duration.

Potential evapotranspiration

Potential evapotranspiration is not measured. It is calculated from other meteorological measured variables. According to the FAO recommendations the daily values of potential evapotranspiration are calculated using Penman-Monteith formula for a reference grass cover. The formula considers the evaporation dependency on the global sunshine irradiance, air temperature, air humidity and wind speed. All the input data for the daily evapotranspiration calculation were adjusted to the last period when the wind speed was measured at automatic weather stations. Monthly value of potential evapotranspiration is the sum of the daily values in a given month.

Air pressure

In the past air pressure was measured at observing stations mostly by a mercury barometer and nowadays it is measured using an electronic device. When the pressure is measured by mercury barometer the observer reads the barometer temperature and the upper level of the mercury column. Both data serve for the calculation of station pressure. The daily mean is the arithmetic mean of measurements at 7, 2 p. m. and 9 p. m. CET.