Search

he rise in temperature in recent years has significantly affected the entire biosphere, including the beginning and duration of the growing season. The earlier start of the growing season is visible in many crops and trees. Although an increase is generally visible in all temperatures (minimum, maximum, and mean daily), this does not exclude the occurrence of extreme weather conditions that can damage crops. In particular, a special danger is related to frost. Since there is no unequivocal method for determination of the occurrence of a frost event, many studies try to determine the risk of its occurrence using different criteria. This study aimed to determine the best method for predicting the frost event in the study area. In that matter, several already introduced methods, were tested on data and observations from meteorological stations all over Croatia. Also, a new approach has been tested, in which frost day is described with the criterion: T min < 3 °C and dew point temperature (calculated using the minimum temperature) T d < 0 °C. It was this method that showed the best results. The probability of detection for this method is higher than 90%, and the probability of false detection was around 15%. Also, all methods were subjected to the Signal Detection Theory. The new approach showed the best results again with a high accuracy index and not showing bias to over or under predicting. Considering all that, this new approach to frost prediction has proven to be very good and provides the possibility provides the ability to predict frost in the future as well as forecast frost a few days ahead which can be significant for the protection of crops.

The rise in temperature in recent years has significantly affected the entire biosphere, including the beginning and duration of the growing season. The earlier start of the growing season is visible in many crops and trees. Although an increase is generally visible in all temperatures (minimum, maximum, and mean daily), this does not exclude the occurrence of extreme weather conditions that can damage crops. In particular, a special danger is related to frost. farmers need information about changes in the number of days with frost, as well as the last days with frost in spring, not only in the current climate but also in the future. For this reason, changes in the number of days with frost in the period from 1981-2020 were observed in this research. In addition, the last day with frost in spring, the first frost day in autumn, as well as the vegetation season in the observed period were calculated. Since there is no simple method for determining a frost day using only measured values and not observations, a new approach was tested, which defines a frost day as a day in which Tmin < 3 °C and the dew point temperature, calculated with Tmin, < 0 °C. This method was more than 90% successful in forecasting frost. Using this method and regional climate models, the change in the occurrence of frost until the end of the century was examined. A decrease in the number of frosty days in the future is visible, but large changes in the occurrence of the last frosty day of spring are not pronounced. If we take into account the expected earlier onset of vegetation, this can be a big problem in the future.

A hail climatology provides knowledge used by other governmental, public, and private institutions for better adaptation to this extreme. Croatia is a small country with a complex topography including the lowland part of eastern Croatia which turns into a hilly part as one goes westward, and the Dinarides rise further south, followed by the Adriatic coast. Thus, diversity of hail climatology throughout the Croatia is analyzed. The results are based on hail data from 199 stations and multiple periods. The results showed significant interannual and spatial variability, due to which three subregions were recognized based on the annual cycle of hail. On the South and central part of Croatian coast, the highest hail activity is present in the colder part of the year while the Northern most parts of the coast and continental part of Croatia shows increased summer hail activity. There is also a transitional area located in between, that records the most hail in spring and fall. The trend analysis was made for three time periods from 1900. to 2020., from 1964. to 2019., and from 1995. to 2019. First two periods have shown negative and significant trend in the number of hail days, while recent, 25 yearlong trend, shows change of sign towards positive but not significant trend. Daily patterns of hail show a shift in the daily maximum from morning to afternoon hours as we progress from the coast towards inland, and the coastal part of Croatia generally records higher hail frequencies compared to the continental part. Most dominant air mass (over 50% of time) responsible for the hail is coming from the SW. Winter hail, is usually induced in stronger sheered environments with lower instability. Climatology of hail duration reveiled a log-normal distribution patter further suggesting that most of the hail lasts between 1 and 5 minutes.

As many studies shows, temperature is key element that affects grapevine growth. Global warming and temperature rise shifted grapevine phenology in many vineyard regions worldwide. In Croatia, viticulture represents one of the most important branches of agriculture in the economical and traditional sense. Therefore, it is important to study changes in vine growth, as well as impact of meteorological parameters on it, so that it can be used to predict suitability and growth under future climate. In this study, dates of beginning of 3 phenological phases (budburst, flowering and veraison), as well as harvest dates, collected from wineries across country, were analyzed. Different statistical methods, that use daily temperature and temperature sum (Growing degree day with base 10 or 5°C) are used to determent beginning of these phases. Results show earlier appearances of almost all phases, regardless of variety. With rise of temperature, the duration between two phases is shorted and that leads to an earlier harvest. Statistical methods generally predict dates well. The best prediction is for budburst and harvest dates with correlation coefficient higher then 0.7. Temperature correlation is worse for predicting flowering and veraison, but shorter period between flowering and budberst as well as between veraison and flowering can correlated with higher temperature.

Hail is a weather phenomenon that can sometimes cause material damage. Although the processes inside the clouds that lead to hail formation are known, there is still insufficient knowledge about this topic. For this reason, numerous researches on hail climatology have been made in the last few years. In spite of the fact that Croatia is a small country, it has a relatively inhomogeneous climate. There can be a lot of reasons for that, for example orography, proximity to the Adriatic Sea etc. Characteristics of hail vary in different climatic conditions, and this research focuses on characteristics of hail in the climactic conditions of costal part of the Adriatic. The results are based on hail data from 55 stations from the observation log for the period from 1973 to 2019, and ERA 5 data for instability indices. The results showed significant interannual as well as spatial variability, due to which it was necessary to make a division into 4 subdomains. Trend analysis showed negative trend in interannual number of hail events, and the loss of hail events was reflected on summer and autumn seasons. It was shown that on the entire coast higher hail activity is present in the colder part of the year, and less activity in the warmer part of the year. Daily patterns showed 3 daily highs – morning, noon and afternoon, and the duration was usually 5 minutes. Three stations were singled out as the ones with the most hail days a year: Drniš, Senj and Gračac. Finally, instability indices were studied (CAPE, KI, DLS and freezing level height), which can be used for explaining the atmospheric conditions in which this extreme phenomenon occurs.

Utjecaj klime na vinogradarsko zoniranje procijenjen je na temelju usporedbe vrijednosti agroklimatskih indeksa poput Huglinovog (heliotermičkog) indeksa, indeksa sume efektivnih temperatura prema Winkleru, indeksa hladnih noći u rujnu i indeksa prosječne temperature zraka u vegetaciji za razdoblje 1988.-2017. u odnosu na referentno klimatsko razdoblje 1961.-1990. kao i linearnim trendovima Winklerova indeksa u razdoblju 1981.-2017. Za sve indekse se je pokazao porast u novijem razdoblju u odnosu na referentno te pozitivan trend Winklerova indeksa posljednjih desetljeća. Iz toga slijedi potreba na cijelom području Hrvatske za promjenama vinogradarskih zona kao i uvođenje C III b zone u podregijama Sjeverna Dalmacija te Srednja i južna Dalmacija.

The factors which significantly affect the viticulture are temperature and precipitation. Consequently, viticulture is highly affected by climate change. The agroclimatic indices describe the suitability of particular region for wine production. For the purpose of this study six indices (Average growing season temperature, Growing degree days, Huglin index, Dryness index, Cool night index and Composite index) were calculated using daily output from three CORDEX Regional Climate Models’ (RCMs) simulations (CLMcom-CCLM4-8-17, SMHI-RCA4, CNRM-ALADIN5.3) for Croatian domain and using daily near-surface measurements (minimum, maximum and mean temperature, relative humidity, wind speed and total precipitation). All RCMs are forced by Global Climate Models (GCMs) with a moderate (RCP4.5) and a high-end (RCP8.5) greenhouse gass (GHG) scenario. All the simulations have horizontal grid spacing of 0.11◦. First historical analysis was done comparing two periods (1961-1990 and 1988-2017) to establish present situation and to determine changes in indices so far. Then, in order to determine future changes in agroclimatic indices, spatial distribution of the indices in historical runs (1971-2000) is compared to three different 30-year periods (2011-2040, 2041-2070 and 2070-2099). The sign and the robustness of future changes depend on the location/region analyzed. The results revealed good skill of the RCM in simulating bioclimatic characteristics. The results also show, whether the part of the country, which is suitable for grape cultivation in present climate (1971-2000) continues to be favorable in the future. We can also reveal whether some other parts of Croatia become suitable for cultivating grapevine in the future climate.

The potential environmental impacts on grapevine yield, sugar and acidity can be very diverse depending on heat and available water in changing climate. Due to the strong dependence of vines on atmospheric conditions, a set of bioclimatic indices has been developed and used in many studies to combine meteorological and grape cultivation information in a quantitative manner. The set mainly consist average temperature growing season (TGS), Growing degree-days (GDD or WI), Huglin Index (HUI), Dryness index (DI), Cool night index (CI). Although many authors calculated these indices for the specific locations, new grided meteorological datasets created from measurements and/or regional climate models (RCMs from EURO-CORDEX) enabled spatial analysis of these indices which allow the viticultural zoning of larger regions. Since changes in temperature and humidity will act (i) differently on all developmental stages of vines in certain wine growing areas (i.e. on their phenological characteristics) and (ii) differently on different varieties, a comprehensive database of bioclimatic indices has been calculated and analysed using all available meteorological and agronomic measurements as well as the results of RCMs. The horizontal grid spacing of 0.11° enabled fine determination of bioclimatic indices for present and future climate in Croatia. In addition, statistical analyses have been done examining sugar content and total acidity and date of harvest in the present and future climate from selected wineries and selected varieties. The results show whether the part of the country, which is suitable for grape cultivation in present climate, continues to be favourable in the future. The spatial distributions in the future climate also suggest that the existing zoning of wine-growing areas is not adequate for the Croatian territory.

Vinogradarstvo i vinarstvo među najvažnijim su poljoprivrednim i gospodarskim djelatnostima u Hrvatskoj. Vinova loza (Vitis vinifera L.) klimatski je osjetljiva kultura, jer na njen rast i razvoj uvelike utječu prevladavajući atmosferski uvjeti. Optimalni klimatski uvjeti za rast vinove loze ograničeni su zemljopisno, te se uobičajeno smatra da su najpogodnija područja za njezin uzgoj između 30° i 50° paralele sjeverne i južne geografske širine. Gledajući kroz sektor vinogradarstva i vinarstva, agroklimatski indeksi su metode kojima se određuju zemljopisna područja i njihova pogodnost za uzgoj određene sorte vinove loze, a pomoću njih se izrađuju karte vinogradarskih područja. Utjecaj klime na vinogradarsko zoniranje procijenjen je na temelju projekcija vrijednosti agroklimatskih indeksa poput Huglinovog indeksa, indeksa sume efektivnih temperatura prema Winkleru, indeksa hladnih noći i indeksa prosječne temperature zraka u vegetaciji. U Hrvatskoj je prva podjela na vinogradarske zone vezana uz izradu regionalizacije vinogradarskih područja početkom 70-tih godina 20. stoljeća. Tadašnja podjela vinogradarskih zona napravljena je sukladno međunarodnim standardima i preporukama OIV-a te Lisabonskom sporazumu iz 1958. godine, a prema kojima su „zone proizvodnje rezultat prirodnih čimbenika i pri čemu klimatske prilike igraju prvenstvenu ulogu“. Prema sumi efektivnih temperatura, današnja vinorodna područja u Europi podijeljena su na klimatske zone: A, B, C I, C II, C IIIa i C IIIb. Ulaskom Hrvatske u Europsku uniju, vinogradarske zone u Hrvatskoj podijeljene su na B, C I i C II zonu. Rezultati najnovijih istraživanja te izračuni modela za projekciju buduće klime ukazuju na prisutnost klimatskih promjena i značajan porast temperature zraka, što podupire hipotezu o nužnosti revidiranja svih vinogradarskih zona u Hrvatskoj.

Monitoring the temperature within the vineyard during the season is important for controling the maturation and planing various production activities in the vineyard. The study displays a description of the spatial variability of the temperature within the vineyard, since it has a strong effect on grape ripening and quality of the grapes. The goal of this study was to collect detailed temperature measurements at the two locations within the vineyard and to estimate temperature sensitivity at the vineyard scale. The results show that there is a great temperature variability within the vineyard and thus the need for system of zonal vineyard management. Rather than being managed uniformly, individual block can be split into zones in which management can be applied differentely.

Climate conditions are one of the most important factors affecting wine, determining whether an area is suitable for planting of a vineyard, and how will express its quality potential.Based on natural conditions, Croatia is divided into viticulture regions: Eastern Continental Croatia, Western Continental Croatia and Coastal Croatia. According to the sum of effective temperatures in the vegetative period, region Coastal Croatia is located in climatic zone CII (1941 to 2220 ºC). According to the official data of Payment Agency for Agriculture, Fisheries and Rural Development, the most important red wine variety in Croatia is Plavac mali, indigenous variety of Croatia with extraordinary quality potential. In this paper, data from the meteorological stations of the Meteorological and Hydrological Service of Agriculture (periods from 1961 to 2014), trends over 30 year periods (1961-1990, 1988-2017) and agroclimatic indices: Growing degree-days (GDD or WI) and Heliothermal index (HI) will be presented.During the harvest in 2016 and 2017, we collected samples of grapes in chosen vineyards throughout the region. Clusters and grape berries were measured and weighed, and by analyzing the must we obtained values of sugar concetration, acidity and pH.The results of this study will be used to evaluate the suitability of potential wine growing areas for cultivation of variety Plavac mali, and to compare it with its current distribution in the region of Coastal Croatia.

The factors which significantly affect the viticulture are temperature and precipitation. Consequently, viticulture is highly affected by climate change. The agroclimatic indices describe the suitability of particular region for wine production, so they are usually used as a tool for viticultural zoning. For the purpose of this study six indices (Average growing season temperature, Growing degree days, Huglin index, Dryness index, Cool night index and Composite index) were calculated using daily output from three CORDEX Regional Climate Models’ (RCMs) simulations (CLMcom-CCLM4-8-17, SMHI-RCA4, CNRM-ALADIN5.3) for Croatian domain. All RCMs are forced by Global Climate Models (GCMs) with a moderate (RCP4.5) and a high-end (RCP8.5) green-house gass (GHG) scenario. SMHI-RCA4 is driven by five different GCMs (CNRM-CERFACS-CNRM-CM5, ICHEC-EC-EARTH, IPSL-IPSL-CM5A-MR, MOHC-HadGEM2-ES and MPI-M-MPI-ESM-LR), CLMcom by four (CNRM-CERFACS-CNRM-CM5, ICHEC-EC-EARTH and MOHC-HadGEM2-ES, MPI-M-MPI-ESM-LR), and CNRM-ALADIN5.3 with one (CNRM-CERFACS-CNRM-CM5). All the simulations have horizontal grid spacing of 0.11°. In order to determine future changes in agroclimatic indices, spatial distribution of the indices in historical runs (1971-2000) is compared to three different 30-year periods (2011-2040, 2041-2070 and 2070- 2100). Also, monthly and seasonal values of predicted meteorological parameters (air temperature, air humidity, precipitation and wind speed) during the growing season (from April to October) have been compared against the corresponding parameters in historical runs. The sign and the robustness of future changes depend on the location/region analyzed. As wine production has a long tradition in Croatia, the results show whether the part of the country, which is suitable for grape cultivation in present climate (1971-2000) continues to be favourable in the future. We can also reveal whether some other parts of Croatia become suitable for cultivating grapevine in the future climate.

Research worldwide indicates that future climate changes will not have an identical impact on all cultivation areas. As wine production has a long tradition in Croatia, it is important to determine whether the part of the country, which is suitable for grape cultivation, continues to be favorable in the future. In order to observe the changes, seven developmental stages of eight phenological stations for the well-known Croatian grapevine varieties (Riesling Italico, Chasselas Blanc, the Queen of Vineyard, Istrian Malvasia, Plavac Mali, Blatina and Trbljan) have been analysed in the period 1961-2017. To indentify the effect of climate change, agroclimatic indices: Huglin heliothermal index (HI) and growing degree-days (GDD), which are frequently used to indicate suitable areas for wine production, have been evaluated in the period 1987-2016 compared to the normal 1961-1990. Linear trend analysis of phenological data showed that dates of leaf unfolding, beginning and end of the flowering occurred earlier by 1-3 days/decade in the Adriatic area and 3-5 days/decade in the continental area. The duration of the ripening period has been reduced by half (on 17-20 days) over the last 30 years. The shortening of the vegetation period is more due to the harvest beginning earlier in the summer than to the earlier onset of vegetation in the spring. In the period 1961-1990 HI was up to 2500°C and GDD up to 2300°C for the mid-Adriatic, up to 2050°C and 1500°C respectively for the continental area and below 1600°C and 1000°C for mountainous area. In the last 30 years, the increase of HI and GDD is evident in the whole Croatia (up to 300°C). Thus, in the near future in the continental part it could become possible to grow thermally more demanding red grape varieties, while earlier varieties could be cultivated in some mountainous parts.

Climate is one of the main factors controlling grapevine production and this relationship is usually explored by the agroclimatic indices. In this study these indices were calculated for Croatian territory from measured and simulated data. First, six indices (i.e. Average growing season temperature, Growing degree days, Huglin index, Dryness index, Cool night index and Composite index) and evapotranspiration by FAO Penman-Monteith method were calculated using daily near-surface measurements (minimum, maximum and mean temperature, relative humidity, wind speed and total precipitation) from numerous land-based stations from 1961 to 2016. Due to better comparison, results are shown for two 30-year periods, 1961-1990 and 1987-2016. The aim is to categorize grapevine growing region climates of Croatia and to see changes in present climate comparing with period 1961-1990. Afterwards, simulated data from three CORDEX Regional Climate Models (RCM) (CLMcom-CCLM4-8-17, SMHI-RCA4, CNRM-ALADIN5.3) at 0.11° grid spacing have been explored. All simulations are forced by the ECMWF ERA-Interim reanalysis for three different periods: 1981-2010 for SMHI-RCA4, 1981-2008 for CNRM-ALADIN5.3 and 1991-2008 for CLMcom-CLM4-8-17. The evaluation has been done for (i) the meteorological variables and (ii) agroclimatic indices using the standard statistical parameters (e.g. bias, RMSD, corelation, etc.). For each approach we show model-measurements in-situ comparison (using the near-neighbors method) and comparison relevant for each region. The second approach was done because of the great variety of climatic characteristics of Croatia. Both approaches revealed good skill of the RCM in simulating bioclimatic characteristics in Croatia. The results also pointed to the categorization of grepevine growing region climates of Croatia and changes in present climate comparing two periods 1961-1990 vs 1987-2016. They are also base for estimating future changes in grapevine growing regions climates of Croatia.

Weather situation during the two forest fires, which were intentionally caused on the Pelješac Peninsula on the night 20/21 July, has been analyzed. The aim of this study is to research weather conditions that led, with the help of human factor to the begging of fire and to validate numerical simulations of models WRF and ALADIN/HR during the first few days of fire. The lack of precipitation and extremely hot weather in July caused the worst possible forest fires - fire crown. During the wildfires slightly lower air pressure prevailed on the southern Adriatic, and the warm air was up to 850 hPa. At the beginning of wildfires air temperature at 2 m achieved 39°C and relative humidity was around 40% during the day. Few hours before fires wind gusts reached 12 m/s. Canadian forest fire weather index (FWI) showed high risk of wildfire until 26 July 2015. Haines index showed high risk of forest fire during the day and moderate risk during nights. Simulated air temperature for both models gave good results, but model WRF produced slightly lower maximum temperature than observed temperature. However, some meteorological elements of the WRF and ALADIN models predicted differently. ALADIN/HR model produced lower relative air humidity, and WRF higher than observations, due to different model setup and selected parameterization. While model ALADIN/HR significantly underestimates 10-minutes mean wind speed, WRF overestimates them. We also compared simulated fields developed by the ALADIN/HR and WRF models with upper air observations. WRF model usually predicts higher wind speed up to 1 km altitude. But it should be noted that weak wind up to 5 m/s prevailed in the troposphere lower layer during the fire except at night on the beginning of fire the WRF model simulated a maximum of 12 m/s. Both models predicted decreasing air temperatures with height and almost neutral stratification of atmosphere. Although the man intentionally set fire, meteorological observations indicate very favorable weather conditions (dry and hot weather) in Pelješac Peninsula in the days before and during fires, for the outbreak and rapid spread of wildfire. Such weather conditions and very steep and inaccessible terrain have hampered fire-fighting which endangered human lives and residential buildings.

In Croatia is observed the increasing trend in mean annual air temperature and the large variability in the measured extreme precipitation from prominent drought to severe flooding. Climatological studies based on regional climate models show the intensification of the extreme values of temperature and precipitation over Europe in the future climate. These changes in temperature and precipitation have a great impact on viticulture, considering that of all living organisms, plants react first to the change of surface temperature. Due to climate change, changes in temperature will affect: (i) differently all developmental stages of vines in certain wine-growing areas in Croatia (i.e. their phenological characteristics) and (ii) differently on different varieties. At the same time, the temperature increase changes convective activity (storms) and thus the occurrence of hail that are estimated as the second biggest cause of material damage in Croatia. The trend of increasing average duration of hailfall was also observed on the basis of measurements obtained in the low land of Croatia. Due to mentioned, the VITCLIC project (VITiculture and CLImate Change in Croatia) has several goals. Based on meteorological and climatological datasets and wine and grapevine archive data, the current situation and a monitoring system of key phenological phases of the vine due to climate change will be analyzed and established. Furthermore, the aim is to link the ripening and quality of grapes with agro-climatic indices and determine the relationship between hail parameters and the degree of damages in grapevines. The purpose is also to enable: (i) detailed agronomic and economic evaluation for preserving Croatian autochthonous grapevine varieties, (ii) determination of climate changes through agro-climatic indices in different future periods and (iii) an estimation of the increase or decrease in certain weather types which favor convective developments (and hail).

In Croatia, forests cover 44% of the area, and the mean annual burned area is about 10, 000 ha mainly on the Adriatic coast and islands. For early warning of fire brigades, it is important to know the weather situation that precedes the wildfires. Two large wildfires were intentionally caused on the Pelješac Peninsula on the southern Croatian coast on 20/21 July 2015 at night and were completely extinguished in mid-August. The aim of this study is to research weather conditions and vertical structure of atmosphere that led to the begging of wildfires by using the meteorological data and ALADIN/HR numerical simulations model. Fires preceded the very dry and hot weather (25 days without rain and seven consecutive days with maximum daily air temperature above 34°C). The absolute maximum temperature of 38.8°C on the peninsula of Pelješac was measured in the first day of the fire, and the next night minimum daily temperature was 30°C. During the wildfires slightly lower air pressure prevailed on the southern Adriatic, the warm air was up to 850 hPa and relative humidity was around 40% during the day. Few hours before fires wind gusts reached 12 m/s. Canadian forest fire weather index (FWI) showed high risk of wildfire until 26 July 2015 and initial spread index (ISI) indicated the worst possible forest fires - fire crown. ALADIN/HR model predicted weak wind up to 5 m/s, decreasing air temperatures with height and almost neutral air stratification in the lower troposphere during the fire. The rapid spread of the fire, which was affected in a short time a large part of the peninsula, favoured very dry dead fuel material of the Mediterranean vegetation and quick lifting of hot air accelerated very steep terrain. Haines index showed high risk of forest fire during the day and moderate risk during nights.

U ovom radu analizirana je vremenska situacija za vrijeme dva šumska požara na poluotoku Pelješcu. Požari su namjerno izazvani u noći s 20. na 21. srpanj 2015. Cilj je rada proučiti vremensku situaciju koja je uz ljudski čimbenik dovela do početka požara. Analizirani meteorološki podaci s najbliže glavne meteorološke postaje Ploče pokazali su da je srpanj 2015. bio najtopliji mjesec u razdoblju 1981.–2014. U srpnju je bilo 30 vrućih dana što je dvostruko više od prosjeka. Izostanak oborine i ekstremno topao srpanj pridonijeli su tome da dođe do najgoreg mogućeg požara, požara krošnji. Produkti modela ALADIN za vrijeme šumskog požara su pokazali malo sniženi tlaka zraka nad južnim Jadranom, a topao zrak se protezao sve do 1 km u vis. Vrijeme je bilo pretežno vedro, na sam dan izbijanja požara izmjerena je u Kuni i apsolutna vrijednost temperature 38.8 °C, a i idućih se dana maksimalna temperatura nije spuštala ispod 30°C na Pelješcu. Relativna vlažnost zraka tijekom dana najčešće je iznosila oko 40%. U noći kada su požari podmetnuti vjetar je bio slab. Indeks meteorološke opasnosti od šumskog požara FWI tijekom svih dana požara bio je vrlo velik, ali to je i očekivano za vrijeme sezone požara. Dugotrajno suho i vrlo vruće vrijeme koje je prethodilo šumskim požarima te vrlo strmi teren poluotoka Pelješca su pogodovali vrlo brzom širenju požara, otežavali gašenje požara što je ugrozilo ljudske živote i stambene objekte.

Ovaj rad bavi se proučavanjem utjecaja meteoroloških uvjeta na trendove početka određenih fenoloških faza vinove loze. U tu svrhu koristili su se osnovni agroklimatološki indeksi GDD i GST. Također, upotrebom statističkih modela simulirali su se nastupi fenoloških faza pupanja, cvatnje, šare i berbe. Statistički pristup kombiniran je i primjenom modela za rast biljaka (model STICS). Sam model STICS testiran je na mjerenjima za dvije sorte vinove loze (’Graševina’ i ’Plavac mali’) i na četiri lokacije; dvije smještene u unutrašnjosti Hrvatske (Daruvar i Križevci) te na dvijema uz obalu Jadrana (Hvar i Lastovo). Uspješnost statističkog modela na temelju akumulirane topline pokazala se vrlo dobrom za sve postaje za berbu kao i za cvatnju, osim postaje Hvar. Lošiji rezultati dobili su se za šaru na gotovo svim postajama te za pupanje izuzev postaje Daruvar. Višestruka linearna regresija pokazala je manju uspješnost. Prilikom mjerenja model STICS pokazao je zadovoljavajuće rezultate na tri od četiri promatrane postaje. Model STICS primijenjen je i na izlazne rezultate klimatskog modela koji je dao rezultate u dva razdoblja: 1999.-2009. godine te 2043.-2053. godine. Navedeni model pokazao se učinkovitim i samim time primjenjivim za simulaciju nastupa fenoloških faza. Uz navedeno, simulirao je i raniji nastup fenofaza u budućoj klimi za gotovo sve postaje. This paper deals with analyzing the influence of meteorological conditions on the starting trends of certain phenological phases of grapevine. For this purpose, the basic agroclimatic indexes GDD and GST were used. In addition, the occurances of different phenological phases, that being budding, flowering, veraisoning and harvesting, were also simulated using statistical models. This statistical approach is also combined with the application of a plant growth model (STICS model). The STICS model itself was tested on measurements for two grapevine varieties (’Graševina’ and ’Plavac mali’) and on four locations; two of which were located in the interior of Croatia (Daruvar and Križevci), while the other two were on the Adriatic coast (Hvar and Lastovo). The success of the statistical model based on accumulated heat was proven to be very good for all harvesting and flowering stations, except for the Hvar station. Worse results were obtained for chara at almost all stations, as well as for budding, with the exception of the Daruvar station. Multiple linear regression showed less success. During the measuring process, the STICS model showed satisfactory results at three of the four observed stations. The STICS model was also applied to the output results of the climate model, which produced results in two periods: 1999-2009. and 2043-2053. The mentioned model was proven to be effective and therefore applicable for simulations of the performance of different phenological phases. In addition, it simulated the earlier occurrence of phenophases in the future climate for almost all stations.