Clilmate Analogies and Risk Analysis of Hungarian Viticulture

IAALD AFITA WCCA2008 WORLD CONFERENCE ON AGRICULTURAL INFORMATION AND IT

Clilmate Analogies and Risk Analysis of Hungarian Viticulture Karoly Szenteleki1, Levente Horvath2 and Márta Ladányi3 1 Corvinus University of Budapest, H-1118 Budapest, Villányi út 29, Hungary,

[email protected] 2 MTA-BCE Adaptation to climate change research group, H-1118 Budapest, Villányi út 29,

Hungary, [email protected] 3 Corvinus University of Budapest, H-1118 Budapest, Villányi út 29, Hungary,

[email protected] Abstract The Department of Mathematics and Informatics of Corvinus University of Budapest can be considered one of the higher level Hungarian climate research centres. Current studies at our department, which have enormous scope—like everywhere else in the world—demand two things: - meteorological databases - large secondary databases (heterogeneous in content and structure and multi-disciplinary: agriculture, health, etc.) To collect, organize, manage and search such databases it was necessary to create a special database system (VIN-MET). It allows researchers and stakeholders to set up special databases for specific studies as well as having the capacity to filter and aggregate data from different perspectives.

Several climate models were elaborated to predict the climate change tendencies. Our target (reference) site was Tokaj, an important centre of viticulture in Hungary. Scenario values for Tokaj were calculating from the TYN SC 1.0 10 minute resolution grid climatic database, in different time periods. Then the method of spatial analogies was used to understand the meaning of these scenario values. Spatial (geographical) analogues are regions which today have a climate analogous to that predicted in the study region in the future. It has been found that regions similar to the predicted future climate of Tokaj are located south to Hungary. At the closest time this distance is about 150-350 km.

Climate change and its expected impacts on viticulture of Hungary are considered and the reasons and consequences of risk increase are explored. As a result of a synthesized analysis of international and national literature we fix some weather indicators which may significantly define grapevine production. Based on RegCM scenarios we introduce the expected change of these weather indicators and formulate some conclusions for Hungary. Keywords: viticulture, climate change, data management, weather indicators, climate

analogies The VIN-MET database

Meteorological and viticultural data are coming from several sources and are to be applied for quite different kinds of purposes. In the planned paper we introduce ‘VIN-MET database management system for viticultural climate change research’ which was developed by the

389

mailto:[email protected]




Research Group of the Hungarian Academy of Sciences for the Adaptation to Climate Change Impacts. VIN-MET is an integrated and extensive database and software which contains detailed meteorological and viticultural data from all vineyard regions in Hungary. We required a system that allows researchers from various disciplines to create special new databases for specific studies as well as having the capacity to filter and aggregate from different

perspectives. Once they have these databases, which are in self-determined format, structure, and size, they can apply advanced mathematical methods and use them as a launching database. Results are collected in Access database format for a flexible form of later applications. The system runs within the department’s internal network. Each researcher logs in with an identification (number) and uses it to create a container (Access database), into which are entered his or her own filtering results in search table format.

In the case of monthly meteorological data, we have available the 10-kilometer rows of data for Europe for 1901-2000, and the meteorological scenarios for 2001-2100 in the same scale. Our research team used the Tyndall database together with free-standing computer database programs. Because data searches require geographical location to be determined, a converting program had to be developed to go between the Tyndall grid and the Hungarian communities. The program is also capable of searching by the regional communities locations as determined by the given geographic coordinates. The system provides precise setting of time periods in addition to search of meteorological data within a restricted area Climatic requirements of agricultural plants The climatic needs of agricultural plants are fairly well documented in the professional literature. It is necessary to have defined climatic conditions for each plant, and this is of course the most important investigation to be carried out by meteorological databases, but more complex mathematical analyses are needed to determine mutual and cross-influences as well as internal relationships.

It is possible to create climatic profile indicators for daily as well as monthly data by combining temperature (minimum, average, maximum) as well as distribution of yearly precipitation (other time periods can be used, also). In the case of daily data, the system of conditions can be set up by day, but for making parameters for longer time periods (weeks, for example), linear interpolation can be applied, and this is done automatically by the computer. This system makes it possible both to define conditions in plant production and assess changes in proliferation of harmful insects. The computer surveys and evaluates past as well as predicted future scenarios to monitor temperature and precipitation characteristics.

Climate Profile Indicator Module supports the search and analysis of some weather indicators for grapevine e.g.: Winkler Index (°C), Huglin Index (°C), Biologically Effective Day Degrees (°C), Mean Temperature of the Warmest Month (°C), Growing Season Maximum Temperature Average (°C), Continentality, Spring Frost Index, Annual Rainfall (mm), Summer

390


Rainfall (mm), Monthly Degree Days (°C), Heat Degree Days (°C), Harvest Maximum Temperature (°C), Number of Rain Days in the Growing Season, Ripening Month Rainfall (mm), etc.. It is capable of implementing the actual daily meteorological data as well as RegCM scenarios.

Applying the so-called Climate Profile Indicator Module Plus we can also create new indicators by defining lower and upper boundary conditions regarding to daily as well as monthly data. It is also possible to combine temperature (minimum, average, maximum), radiation as well as precipitation data of any time period. In the case of daily data, the system of conditions can be set up by day, while for making parameters for longer time periods (weeks, for example), linear interpolation can be applied in a very

simple and user friendly way. The module makes it possible e.g. to define conditions for production quantity or quality demands or help to assess the risk of changes in occurrence and spread of pests or diseases. The software can survey and evaluate historical data and RegCM scenarios to monitor temperature, precipitation and radiation characteristics and helps to decide whether the examined variables (of the examined time series) indicate sufficient or not sufficient conditions, according to the profile indicators. VIN-MET is suitable also for finding the co-existence or absence of more than one meteorological profile indicators at the same time.

Climate analogies based on VIN-MET Baseline climate data of Debrecen were taken from the OMSZ (Hungarian Meteorological Office) database. For the European base climatology (1961-1990) the TYN CL 2.0 [1] database was used, which contains 10 minutes spatial resolution grid data for 5 climatic parameters. From these only the monthly temperature and precipitation data were used. We used the well-known, internationally used emission scenarios – A1, A2, B1, B2 – from the Hadley Centre’s HadCM3 model. In this paper the results of the A1FI scenario are presented. For the scenarios, the TYN SC 1.0 [2] database was used, and we calculated the monthly averages for the 2011-2040 time period.

For the calculation of spatial analogy, two methods were used, the Rank-method and the CLIMEX method. With the Rank-method for each grid boxes the Euclidean distances of the climatic parameters were calculated and ranked. This method shows only the closest points, the most similar grid’s similarity is 100 the less similarity is 0. That means that we always have 100% similar point, which maybe not really similar. The other disadvantage of the method is that the similarity depends on the number of the examined grid points. With the CLIMEX method the similarity index for two climate datasets can be given so, that with parameter settings the user can define, what he calls similar. The method’s advantages are that does not depend on the number of grid points, and shows really the similarity.

391


Rank-method 2011-2040 CLIMEX method

Fig. 1. Spatial analogues for the A1FI scenario For proofing that the methods are working well, the analogue regions for the base period were calculated. The method is appropriate if the spatial analogues of the examined region (Debrecen) are also in the examined region. In the maps (Fig.1.) can be seen that the analogue regions are mostly at the surroundings of Debrecen. The Rank-method shows higher similarity than the CLIMEX method, this is because in case of the ranking, 10% of the grid points always have more than 90% similarity.

1961-1990 Analogy using the temperature (λ=1)

Analogy using the precipitation (λ=0)

2011-2040

Fig. 2. Analogue regions for the base and 2011-40 by the temperature and the precipitation

The difference of the methods can be seen at the calculation for the scenarios, too. They

show similar spatial analogues, but the Rank-method gives greater analogue regions, and shows higher similarities. With this method, probably we get back also such regions, which have not really similar climate. For this reason the CLIMEX method was selected for further analyses.

The analogue regions obtained by the temperature (λ=1) and precipitation (λ=0) can be seen in Fig.2. Results show similarities with the previous maps (temperature and precipitation together ), in the base period give back the surroundings of Debrecen, but other areas also appear as analogues. The same tendency can be observed in case of the scenario. These regions can be analogues regarding temperature but not precipitation, or vice versa.

392


But if the purpose of the analyses does not require the use of only one climatic parameter, we recommend using the CLIMEX method with the temperature and precipitation together, with the same weights. But if the purpose of the analyses does not require the use of only one climatic parameter, we recommend using the CLIMEX method with the temperature and precipitation together, with the same weights. We found that the method of the spatial analogy is a good tool understanding the effects of the climate change. Finding the analogue regions we can compare their cropping system, land use and other ecological and economic characteristics with the present ones in Hungary. It helps developing adaptation and mitigation strategies on climate change. Risk analysis based on VIN-MET There were available historical and control weather data for the baseline periods 1961-90 or 1901-2000. Moreover, there were available RegCM scenarios with reference period 2071-2100. For the period in between, however, we needed a weather generator the weather data of which can somehow connect the two periods. We needed it because stakeholders can be persuaded only if we draw a possible scheme of near future as well. To this, we used C2W which is aimed at disaggregating climatological means and anomalies into realistic weather processes.

Biologically Effective Day Degrees (°C) of Tokaj and Sopron are also both expected to increase if we compare the values based on time series 1901-2000 and 1991-2070 (A2 and B2 scenarios). Moreover, not only variances but also coefficients of variances (%) increase significantly, too.

Winkler Index is: the sum of average temperature experienced above the baseline (10°C) in the entire growing season (April to October). Biologically Effective Day Degrees (°C) is derived from Winkler index with a cut-off at 19°C (Gladstones, 1992). According to Riou’s (1994) Winkler Index classification of the viticultural climatic regions, Sopron and Tokaj grapevine production regions of Hungary are belonging to the first (below

1390 °C) and the second (1391-1670 °C) coolest classes based on temperature data of 1901-2000. Both regions, however, are expected to be belonging to the fifth (warmest) regions (over 2220 °C) up to 2070 based on A2 and B2 scenarios’ data of Prudence Hadley Center (Figure 3). Compare this with the fact that the fifth class corresponds to the current Winkler Index values of e.g. Split (South-Croatia), Palermo (Italy) or Algiers (Algeria).

393


800900

100011001200130014001500160017001800

1880 1900 1920 1940 1960 1980 2000 2020

°C

year

Winkler_Tokaj_1901-2000

800

900

1000

1100

1200

1300

1400

1500

1600

1880 1900 1920 1940 1960 1980 2000 2020

°C

year

Winkler_Sopron_1901-2000

800

1300

1800

2300

2800

3300

1980 2000 2020 2040 2060 2080

°C

year

WTokCRU_A2

800

1300

1800

2300

2800

3300

1980 2000 2020 2040 2060 2080

°C

year

WSopCRU_A2

Figure 3 Winkler index (Tokaj and Sopron regions) calculated for 1901-2000 (1.) as well as for 1991- 2070 (scenarios). The weather data were interpolated by C2W weather generator between CRU and Prudence Hadley Center (HC) A2.

Huglin Index (°C) provides us some more information than Winkler Index does as daily maximum is considered, too. According, we can conclude that up to 2070 Hungarian grapevine growing regions may shift from temperate class (over 1800°C and under 2100 °C) to warm class (over 2400 °C and under 3000 °C). It means that, supposed that humidity, radiation and wind circumstances allow it, regions producing now Riesling, Pinot Noir, Chardonnay, Merlot or Cabernet franc may become to be suitable for Cabernet Sauvignon, Grenache or even all cultivated varieties as there is no more constraint to ripen them. Till the end of the century, however, the values of Huglin Index may exceed the needs of even the late varieties resulting high risk of production.

The grapevine ripening capacity of a region has also been related to the mean temperature of the warmest month or, on the north hemisphere, shortly to Mean July Temperature (°C) Mean July Temperature (°C) indices calculated for the scenario data are significantly higher than the ones calculated for time series 1901-2000 in both Tokaj and Sopron regions. Note that not only variances but also coefficients of variances (%) are much higher for scenario data which means us significantly hotter and more variable July temperatures in the next few decades. The rate of increase is about 4°C. Harvest Maximum Temperature (°C) and Winter Minimum Temperature (°C) are indices with which we can filter years with extreme high and low temperature events that can result serious damages. Winter Minimum Temperature, moreover, can be applied to the risk analysis of pests and diseases the lifecycle, spread and/or the rate of invasion of which are expected to change in Hungary with high winter minimums.

394


Applying Growing Season Length we can show how present site classification is changing with climate change since because of warmer springs and thus earlier bud break and flowering, the growing season length increases in Hungary. The Growing Season Maximum Temperature Average (°C) index is created from the average monthly maximum temperature of the pre-harvest months (June-September). The index is particularly important from the

aspect of both quality and quantity of production. The increase of the averages, the variances and the coefficients of variances for both regions is almost threatening.

Grapevine is quite sensitive to both precipitation and humidity. Yet inn the last 10-20 years vineyards in Hungary have been suffering from relatively arid, hot and low humidity summer climate. Though there is no agreement if Annual Rainfall (mm) in Hungary is going to decrease at all, the yearly distribution and variety of it are expected to change considerably). Summer Rainfall (mm) is very important as besides warming, precipitation in expected to decrease in summer significantly, during the growing period of grapevine. Though there is significant provable change neither in averages, nor in variances or coefficients of variances (%) of the Annual Rainfall (mm) indices of Tokaj and Sopron, the case of Summer Rainfall (mm) indices is quite different. We can detect significant decrease in the averages and, meantime, an increase in variances and coefficients of variances (%). It means that climate change brings us not only much hotter but also much drier summers to Hungary. The Number of Growing Season Rain Days gives some information about the (unbalanced) distribution of future precipitation. Ripening Month Rainfall (mm) is the base of berry splitting and emergency early harvest events’ risk assessment. Conclusions The two examined methods – Rank-methods and CLIMEX – for finding the spatial analogue regions give almost the same results. But we recommend using the CLIMEX method with the same weights of temperature and precipitation. If we accept the results of the GCMs, according to the A1FI scenario for the 2011-2040 periods, the analogue regions of Debrecen will be at Vojvodina region in Serbia, and South-Romania. It means about 100-300 km shifting to south, which correspond to other international results, but later the scenarios predict more drastic change, the climate shifting will be 400-600 km to South. The detailed analyses of the analogue regions can help us to adapt to the changing climate. From the analogue regions we should collect all kind of available ecological, agricultural, economical, social or public sanitation data. We can study what kind of problems are there, and what are the solutions. We can learn from there how to solve the possible problems and develop strategies. This will be a good base for further research and an important base for decision makers.

395


According to the downscaled Prudence A2/B2 scenarios we can expect the following changes in Hungary up to and after 2070: -Increasing mean temperature. Reduced winter dormancy. Warmer and longer growing seasons and season’s shift of 6-25 days earlier over numerous varieties and locations. Earlier ripening. Higher temperature at ripening. Locations of varieties common in Hungary are expected to be shifted northward, while current grape growing regions may shift into another maturity type. Increasing winter-spring and decreasing summer-autumn precipitation. Less water is available, increased yield/quality variability, higher economic risk for the producer. More frequent and serious anomalies. More complex management is needed. Irrigation is needed. -Warming during also the dormant periods. Thus the number of winter/frost days decreases, nevertheless, late frost risk increases. Longer extreme periods as well as increased seasonal variability. Risk increase in widely distributed sectors in connection with grapevine production. Changes in the presence or intensity of pests and diseases. New technologies and management practices are needed. -Now it is obvious that Hungarian grapevine production depends highly on the magnitude, rate and distribution of future warming, precipitation and extreme events. We can recognize that for grapevine production we need the assessment of the expected impacts and it is necessary adapt to them accordingly by altering varieties and management practices or mitigate production quantity and/or quality risk by developing new technologies in time. Acknowledgement Our work was supported by NKFP-B3-2006-0014 References Gladstones, J. S. (2000) Past and Future Climatic Indices for Viticulture. 5th International

Symposium for Cool Climate Viticulture and Oenology, 2000, Melbourne, Australia. Herdon M., Kovács Gy., Magó Zs. (2006): Számítógép-hálózatok HEFOP elektronikus jegyzet.

DE ATC AVK, Debrecen, 2006. 1-131 p. Horváth, L. Gaál, M., Solymosi, N. (2008) Use of Spatial Analogy to Understand the Effects of

Climate Change. International Symposium (Interdisciplinary Regional Research) ISIRR 2007 Novi Sad

T.D. Mitchell, T.R. Carter, P.H. Jones, M. Hulme, M. New, “A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: the observed record (1901-2000) and 16 scenarios (2001-2100)”, Journal of Climate, 2003. Tyndall Centre Working Paper 55, University of East Anglia, Norwich, UK, http://www.tyndall.ac.uk/publications/working_papers/wp55.pdf

Solymosi, N., Kern, A., Maróti-Agócs, Á., Horváth, L., Erdélyi, K. (2008). An easy to use tool for extracting climatic parameters from Tyndall datasets. Environmental Modelling & Software. (in press)

A.M. Young, B. Blackshaw, G.F. Maywald, R.W. Sutherst, R.W., “CLIMEX for Windows 1.1. Tutorials”, CSIRO, Melbourne, 1999.

Winkler, A. J., Cook, J. A., Kliewer, W. M. and Lider, L. A. (1974) General Viticulture . p143-144. University of California Press, Berkeley.

396

http://www.tyndall.ac.uk/publications/working_papers/wp55.pdf

Documents

Clilmate Analogies and Risk Analysis of Hungarian Viticulture