The Impact of Temperature and Precipitation Change on the Production of Grapes in the Czech Republic

Abstract

The warming of the planet and ongoing climate change are now a scientifically proven fact. These phenomena have an impact on nature and many human activities, but logically affect agriculture the most. History has confirmed that the production of grapes (the extent and quality) is significantly affected by climate change. The main goal of this study was to evaluate the impact of climate change through changes in average precipitation and average temperatures on the quantity of grape production in the Czech Republic. A partial goal was then to predict the future development of grape production depending on the expected total precipitation and average temperatures. The effect of changes in average temperatures and total precipitation was evaluated using multiple linear regression methods. The multiple regression model did not reveal a dependence of the total precipitation and average temperatures on the development of the value of vine production due to the statistical insignificance of the effect of average temperatures on the value of vine production. However, when abstracting the effect of average temperatures on the value of vine production, the research confirmed the effect of the change in total precipitation on the value of vine production. The analysis identified the effect of changes in total precipitation and temperatures on the production of grapes in the Czech Republic.

1. Introduction

Until practically the middle of the 20th century, the opinion that the climate was not subject to change (even among the professional public) prevailed across the world. Opinions then changed gradually due to rising temperatures after 1880 [1]. The monitoring of the climate and its fluctuations (especially the monitoring of the development of temperatures and precipitation) in the 20th century also contributed positively to the change in the approach to climate change assessment. Until then, data that could be analyzed were practically unavailable. In particular, temperature fluctuations were monitored, which were often accompanied by significant negative economic impacts [2]. Climate change is generally defined as a development in the climate that has been ongoing unilaterally over a period of time (warming/cooling is especially monitored).

Only during the 20th century, when precipitation and temperature data were collected regularly, did the average temperature in Europe increase by 1.2 °C and tropical days double [3]. Meteorological measurements and monitoring also show that the average air temperature in the Czech Republic has been increasing for a long time, and the incidence of extreme climatic events has also increased [4]. All of this is reflected in agricultural and vine production. This fact also has an impact on the structure of cultivated agricultural crops in the Czech Republic, as evidenced by, e.g., Trnka et al. [5].

The cultivation of grapes is influenced by a number of factors, among which climate change (and especially the change in temperature and precipitation) can be considered a key factor due to the high specificity of the plant’s demands regarding soil and climatic conditions [6,7]. The history of vine growing clearly shows that climate change has a significant impact on grape growing. Grapes are highly susceptible to both short-term and long-term climate change [8]. The growing season of vines is about 180 days, and the plant requires a stable climate without large temperature fluctuations for high-quality growth [9]. Thus, the impact of different environmental factors on production varies considerably; what they have in common, however, is the impact of climate change on the quality and extent of grape production. These changes affect the composition of the grapes, their sugar concentration, and their acidity. For example, Penning-Rowsell [10] and Unwin [11] pointed out that the weather and climate have a key influence on the production and, in particular, the quality of the wine.

In general, the expected climate change during the grape ripening period can be divided into two scenarios: warmer and drier and warmer and wetter, with different responses to red and white grape varieties. Jones [12] has shown that the impact of climate change varies by variety. European regions currently have optimal temperatures; in the future (the next 50 years), an average warming of two degrees Celsius is expected. For regions that now produce high-quality grapes at the margins of their climate limits, these results suggest that future climate change will exceed the climate threshold, making the maturing of grapes for existing wine varieties and styles progressively more difficult.

Other projections have shown that warming may range from 0.8 to 6.6 °C from 2020 to 2080, with precipitation expected to decrease by up to 7–22% during the growing season. The observed temperature trends and expected future climate change directly affect the viability of agricultural production [13,14,15,16,17], leading to the need for ongoing research and understanding to reduce the sensitivity of the wine industry to expected climate change [18,19].

In general, the types of grapes that are grown and the overall style of vines produced in a given region are the result of the underlying climate, while climate variability determines the differences in quality between vintages and vineyards [20]. Understanding climate change and its potential impacts on natural and human systems is becoming increasingly important as greenhouse gas levels, average temperatures, precipitation, and other characteristics of the earth’s environment change. These are then very much reflected in agricultural production and its performance, profitability, and efficiency [21]. There is an increase in average temperatures, but also a higher temperature volatility [22]. Increases in the rate of evaporation of atmospheric water vapor may contribute to the increases in warming [23,24,25].

The main goal of this article is to evaluate the impact of climate change through changes in average precipitation and average temperatures on the size of grape production in the Czech Republic. A partial goal is then to predict the future development of grape production depending on the expected total precipitation and average temperatures.

2. Materials and Methods

The data used in this article are based on data from the Czech Hydrometeorological Institute [26] and the Czech Statistical Office [27]. In the first phase, the development of grape production in the Czech Republic was evaluated using an analysis of time series, and the trend of value development of grape production was analyzed in thousands of CZK. The production of grapes was analyzed with regard to the availability of data from 1990 to 2019. Time series of the value of agricultural production (grapes) were used and were expressed in thousands of CZK. Constant prices were used to eliminate the effect of inflation, i.e., the value of production was adjusted for the effect of inflation and thus expressed exclusively the production of grapes. The constant prices of 1989 were used here. Elementary characteristics for the description of the time series were used for the basic analysis of the time series [28]. In the following key phase of the research, an analysis of the impact of climate change on vine production was carried out. The impact of climate change on the value of vine production was assessed using the effect of changes in average temperatures and total precipitation in the Czech Republic. Both the annual values of the development of average temperatures and total precipitation and the values for the growing season were evaluated. The calculations and presented results were realized using SPSS and Eviews. The Dickey–Fuller test was used to examine the stationarity of the time series. The hypothesis tested was the non-stationarity hypothesis, where t-statistics do not have a standard t-distribution, but a distribution simulated by Dickey and Fuller [29].

The test criterion is given by the relationship:

$$t=\frac{\Phi -1}{S_{\Phi }}$$

(1)

where

• t is the test criterion,
• $\Phi$ process root,
• $S_{\Phi }$ estimate of the standard error of the estimate $\Phi$.

Additive seasonality decomposition was used in cases of significant seasonality of the time series. A key part of this research deals with the analysis of the influence of average temperatures and precipitation on the production of grapes in the Czech Republic. For this, multiple linear regression was first used [30,31] with a dependent variable production of grapes and independent variable average temperatures and total precipitation. The quality of the regression model was evaluated using the coefficient of determination. The value of the coefficient of determination lies in the interval from 0 to 1. The coefficient of determination is given by the relationship:

$$I^2=\frac{S_T}{S_y}$$

(2)

where

• I² coefficient of determination,
• $S_T$ theoretical sum of squares,
• $S_y$ total squares.

The autocorrelation of the time series residuals was determined using the Durbin–Watson autocorrelation coefficient. Based on the values of the Durbin–Watson coefficient, it is possible to ascertain the presence of a first-order autocorrelation, outside the intervals of inconclusiveness. The coefficient has a symmetrical distribution with the mean value E(d) = 2. Values close to the mean value of E(d) represent the series independence of a random component. The DW test is given by the relationship:

$$DW=\frac{\sum {\left({\widehat{\epsilon }}_t-{\widehat{\epsilon }}_{t-1}\right)}^2}{\sum {\widehat{\epsilon }}_t^2}\in \langle 0,4\rangle$$

(3)

where

• ${\widehat{\epsilon }}_t$ residuals at time t,
• ${\widehat{\epsilon }}_{t-1}$ residuals at time t − 1.

The statistical significance of the determined individual regression coefficients was evaluated using t-tests, which have a Student’s distribution of (n − p) degrees of freedom.

$$T=\frac{{\widehat{\beta }}_j}{s_{{\widehat{\beta }}_j}};T~t\left(n-p\right)$$

(4)

where

• $\widehat{\beta }$ regression coefficient,
• $s_{{\widehat{\beta }}_j}$ regression coefficient estimation error,
• n number of observations,
• p number of regression parameters.

The statistical significance of the regression model was determined using the F-test. The F-test used in the model follows Fisher’s distribution.

$$F=\frac{\frac{S_T}{p-1}}{\frac{S_R}{n-p}};F~\left(p-1,n-p\right)$$

(5)

where

• $S_T$ theoretical sum of squares,
• $p$ number of regression parameters,
• $S_R$ residual sum of squares,
• n number of observations.

The Shapiro–Wilk normality test [32] was used to test the normality of residuals. The maximum value of the test is 1, and this indicates a normal distribution. The Shapiro–Wilk normality test is given by the relationship:

$$W=\frac{{\left({\sum }_{i-1}^n{a}_i{x}_{\left(i\right)}\right)}^2}{{\sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2}$$

(6)

where

• $x_{\left(i\right)}$ the i-th value of the ordered vector,
• $a_i$ weight,
• $x_i$ values of the variable,
• $\overline{x}$ average value of values x_i.

The Breusch–Pagan heteroskedasticity test was used for the homoskedasticity test. The simple exponential smoothing method, which was selected due to the short-term constant trend and the ARIMA model [33], was used to estimate the development of grape production. The RMSE coefficient (Root-Mean-Square Error) was used to compare individual models. The square root of the mean square error is given by the relationship:

$$RMSE=\sqrt{\frac{\sum {\left(y_t-{\widehat{y}}_t\right)}^2}{n}}$$

(7)

where

• $y_t$ time series values,
• ${\widehat{y}}_t$ smoothed time series values,
• $n$ number of observations.

The relative forecast error was used to test the quality of the prediction. The value of this error is expressed as a percentage.

$$r=\frac{\left|{y_i}^{\prime }-y_i\right|}{y_i}\ast 100$$

(8)

where

• $y_i\prime$ estimated value,
• $y_i$ actual value.

3. Results

The research was divided into several phases. First, the development of grape production and the development of average temperatures and total precipitation in the Czech Republic were analyzed. The following phase analyzed the effect of changes in average temperatures and average precipitation on the production of grapes in the Czech Republic.

3.1. Development of the Value of Grape Production, Average Temperatures, and Precipitation in the Czech Republic

The basic analysis of the development of grape production was evaluated by analyzing the time series of grape production, the source of which was the Czech Statistical Office. Data from 1990 to 2019 were used. In addition, in these years, the data were used in multiple linear regression, which was then used to examine the dependence of grape production (in thousands of CZK) on the average air temperature [°C] for individual years and the total precipitation (in mm).

Figure 1 shows the development of grape production in thousands of CZK (at constant prices) in 1990–2019. The lowest value of production was recorded in 1997, namely, 286,064 CZK. On the contrary, the highest value in the period observed was recorded in 2018, of 829,633 CZK. Geographically, the viticulture in the Czech Republic is located in two main wine-growing areas, namely, the wine-growing region of Bohemia and the wine-growing region of Moravia. In the Moravia region, vine growing is a historically established practice, mainly due to the very favorable climatic conditions and high soil quality, which are suitable for vine growing [34]. More than 90% of the total vineyard area in the Czech Republic are located in this region [35]. From the development of the time series, it was not possible to recognize whether the statistical moments changed over time on the basis of the graphical display and, thus, whether the probability function of the process was time dependent. At the same time, it was possible to state a generally growing trend over time for the observed period. The Dickey–Fuller test was used to test for the presence of a unit root.

The Dickey–Fuller test hypothesis is defined as follows:

0: 1 = 1,

1: |1| < 1.

The tested hypothesis expresses a situation where the process contains a unit root and the time series is non-stationary, while the alternative hypothesis states that the process has a stationary form. The test results are shown in

Table 1. Dickey–Fuller test of the stationarity of grape production in the Czech Republic.

		t-Statistic	Prob.
Augmented Dickey–Fuller test statistic		−4.626165	0.0009
Test critical values	1% level	−3.679322
	5% level	−2.967767
	10% level	−2.622989

.

The test statistics result was −4.626, and the p-value of the test was 0.0009. The p-value was significantly lower than the standard 5% level of significance, and therefore, we rejected the null hypothesis of the non-stationary time series at the selected level of significance and we accepted the alternative hypothesis, i.e., without a trend. The time series of grape production in the years 1990–2019 can be considered stationary.

The same procedure was applied to the analysis of total precipitation and average temperatures. Within these variables, seasonality was also examined, which plays a key role in the further analysis, especially with regard to the selection of average annual values and the sum of values only in the growing season, which is considered important for the production of vines.

Figure 2 shows the development of the monthly precipitation totals from the year 1990 to 2019. It can be seen from Figure 2 that this is a time series with more significant fluctuations, with the highest total precipitation occurring in July 1997 and August 2002. Given the fact that the grapes grow in the growing season from April to October, it is appropriate to examine the possible seasonality in terms of monthly development and, if seasonality is demonstrated, to focus on the total precipitation during the growing season. Before testing for the presence of seasonality, the Dickey–Fuller test was used to test the stationarity of the time series (

Table 2. Dickey–Fuller test of the stationarity of total precipitation in the Czech Republic.

		t-Statistic	Prob.
Augmented Dickey–Fuller test statistic		−4.490468	0.0002
Test critical values	1% level	−3.448998
	5% level	−2.869653
	10% level	−2.571161

). Based on the result of the stationarity test, we decided whether to use additive or multiplicative decomposition.

The value of the tested criterion was −4.4905, and the p-value was 0.0002. At the 5% level of significance, the null hypothesis was rejected, and an alternative hypothesis of time series stationarity was accepted. Given the fact that it is a stationary time series, additive seasonality decomposition was used, and individual seasonal fluctuations were described using seasonal variations, which were defined as the difference between the time series value and the value adjusted for seasonal fluctuations.

The results (see

Table 3. Seasonal variations in total precipitation in the Czech Republic.

Months	Seasonal Variations (mm)	Months	Seasonal Variations (mm)
1	−11.94113	7	31.77774
2	−21.52734	8	18.79163
3	−10.1624	9	2.90657
4	−17.32044	10	−9.06182
5	13.35772	11	−9.6624
6	23.02151	12	−10.17964

) clearly show that the largest fluctuations in total precipitation occurred in February, June, and July. Given the values of seasonal variations and their differences, the total precipitation in the growing season of the vine was used for further analysis. This allowed us to achieve more accurate results than using the annual total precipitation.

Figure 3 shows the development of the monthly data for average temperatures in degrees Celsius between 1990 and 2019. Temperatures were significantly more regular than precipitation fluctuations, and this regularity can be attributed to the seasonal component. Again, in addition to the graphical evaluation of seasonality, the Dickey–Fuller test was used to test the stationarity of the time series (

Table 4. Dickey–Fuller test of the stationarity of total precipitation in the Czech Republic.

		t-Statistic	Prob.
Augmented Dickey–Fuller test statistic		−4.027793	0.0014
Test critical values	1% level	−3.448889
	5% level	−2.869605
	10% level	−2.571135

).

Based on the result of the test statistics, which was −4.0278 with a p-value of 0.0014, the null hypothesis of the non-stationarity of the monthly time series of temperature development in the years 1990 to 2019 was rejected. The alternative hypothesis was accepted at the 5% level of significance, and we continued to work with the time series based on stationary. Given the fact that it was again a stationary time series, additive seasonality decomposition was used, and individual fluctuations were described using the seasonal variations (

Table 5. Seasonal variations of average monthly temperatures in the Czech Republic.

Months	Seasonal Variations (°C)	Months	Seasonal Variations (°C)
1	−9.75314	7	9.96039
2	−8.78733	8	9.57761
3	−5.10745	9	4.6618
4	0.15433	10	−0.15429
5	4.85491	11	−4.83159
6	8.25232	12	−8.82756

).

The results of seasonal variations show that seasonality was most pronounced in January, July, and August. Given the significant temperature differences, average variations in the growing season of the vine were selected for further analysis. The average value from the given months was calculated as an arithmetic average due to the fact that it is an interval time series. The development of total precipitation (mm) and average temperatures (°C) from the grape-growing months in the years 1990 to 2019 can be seen in Figure 4. The development of average temperatures during the growing season is shown in black, and the development of total precipitation during the growing season is shown in gray.

3.2. Influence of Changes in Average Temperatures and Precipitation on the Production of Grapes in the Czech Republic

Multiple regression analysis was used to analyze the impact of climate change (or changes in average temperatures and total precipitation) on grape production (see

Table 6. Results of multiple linear regression: correlation and Durbin–Watson.

Model 1	Correlation Coefficient	Coefficient of Determination	Adjusted Coefficient of Determination	Durbin–Watson Test
1	0.580 2	0.337	0.288	1.974

¹ Dependent variable: grape production in thousands of CZK at constant prices, 1989. ² Predictor: (constant), precipitation total (mm), temperature (°C).

and

Table 7. Results of multiple linear regression: mean error and p-value.

Model 1		B	Mean Error	Beta	t	p-Value
1	Constant	229,789.573	485,928.332		0.473	0.64
	temperature (°C)	48,740.267	31,019.416	0.264	1.571	0.128
	precipitation total (mm)	−735.541	286.525	−0.431	−2.567	0.016

¹ Dependent variable: grape production in thousands of CZK at constant prices, 1989.

). Within multiple linear regression, the production of grapes (in thousands of CZK) was selected as the dependent variable, and the total precipitation and average temperatures were the independent variables.

The value of the coefficient of determination was 0.337, so the model was able to explain 33.7% of the variability in grape production. The value of the Durbin–Watson coefficient was 1.974; this value indicates the absence of an autocorrelation. The resulting values of the constant and regression coefficients are provided in the output above. The significance of individual coefficients was evaluated using t-tests, the values of which are also given in Table 7. Significance was also assessed using the above p-values, where it can be seen that the p-value of the constant and temperature was statistically insignificant at the 5% level of significance. This means that the temperature variable should be excluded from the model, and we should further work only with the value of the total precipitation, which was significant according to the p-value of 0.016.

3.3. Linear Regression of the Dependence of Grape Production on Total Precipitation

The newly estimated model, where the explained variable is the production of grapes and the explanatory variable is the total precipitation [mm], is as follows:

$$Y=968,090,443-896,976x$$

(9)

The coefficient of determination was 0.276, i.e., 27.6% of the variability of grape production in thousands of CZK was explained by the total precipitation (

Table 8. Results of a linear regression of the dependence of grape production on total precipitation: correlation and Durbin–Watson.

Model 1	Correlation Coefficient	Coefficient of Determination	Adjusted Coefficient of Determination	Durbin–Watson Test
1	0.526 2	0.276	0.250	1.793

¹ Dependent variable: grape production in thousands CZK at constant prices, 1989. ² Predictor: (constant), precipitation total (mm), temperature (°C).

and

Table 9. Results of a linear regression of the dependence of grape production on total precipitation: mean error and p-value.

Model 1		B	Mean Error	Beta	t	p-Value
1	Constant	968,090.443	127,091.564		7.617	0
	precipitation total (mm)	−896.976	274.398	−0.526	−3.269	0.003

¹ Dependent variable: grape production in thousands CZK at constant prices, 1989.

). The value of the correlation coefficient was 0.526, which indicates the mean strength of the indirect linear dependence. The F-test value in the model was 10.686, and the p-value of the F-test was 0.003, which means that the model was statistically significant. At the same time, the results of the p-values of the individual t-tests on the regression coefficients show that the constant and the direction of the regression line were statistically significant in this model.

The results (see Table 9) show that if the total precipitation for the growing season of the vine increases by 1 mm, then the production of grapes will decrease by an average of 896,976 CZK. The relationship between grape production and total precipitation (mm) can be seen in Figure 5.

The selected regression model showed the following values:

$E\left(e_i\right)=0$: random error has a zero mean value;

$D\left(e_i\right)={\sigma }^2<\infty$: finite and constant variance (homoskedasticity);

$Cov\left(e_i{e}_j\right)=0$: zero covariance of a random component for each i ≠ j, where i, j = 1, 2, … n,

$e_i~N\left(\sigma ;{\sigma }^2\right)$: normal distribution of a random component.

Figure 6, Figure 7 and Figure 8 show the results of the regression model, namely, the histogram, residuals, and regression standardized residuals. An evaluation of this model is provided below:

From the resulting graphs above, it can be seen that according to the histogram, the data have an approximately normal distribution. This fact was also confirmed by the value of the Shapiro–Wilk normality test, which was 0.975, and the p-value was 0.6700. The p-value was greater than the standard 5% significance level. We did not reject the null hypothesis of a normal distribution of residuals at the 5% level of significance. The normal distribution was further confirmed by the P-P plot. The autocorrelation of a random component had already been evaluated using the Durbin–Watson coefficient, which was 1.793. The Breusch–Pagan homoskedasticity test provided a value of 0.003585, and the p-value of the test was 0.9527; therefore, we did not reject the null hypothesis of homoskedasticity at the 5% level of significance. The assumption was met. The mean residual value was 0.000, and this is also evident from the Figure 8 of the standardized residuals and values. All the assumptions of the model were thus met.

3.4. Estimated Development of Grape Production for 2020

The last phase of the research was focused on predicting the development of grape production. Given the stationary form of the time series, and thus an insignificant trend, it was not possible to use the trend functions for prediction, so the simple exponential smoothing method was selected to estimate the development, which is suitable if the trend can be considered constant in short sections, which was the case for the time series of grape production in the years 1990 to 2019. This method uses all the known observations of the time series and, at the same time, determines the most current observations in the time series as the most important. The choice of the alpha smoothing constant is made according to the RMSE (Root-Mean-Square error) criterion, which is used to decide on the most appropriate exponential model. For our selected model, alpha = 0.136 was evaluated as the best value. The alpha value was small, which means that the effect of past observations fades very quickly. Higher RMSE values are obtained for higher values of the smoothing constant, which indicate worse models. The RMSE value for alpha = 0.136 was 129,345 CZK. The value of the RMSE error was relatively high, so it was necessary to take this value of the estimate as an approximate estimate and further strive for a prediction depending on the weather and the total precipitation, which affect the development of the time series. The resulting model can be seen in Figure 9.

According to the simple exponential smoothing model, the estimated value for 2020 was 611,382 CZK. At the time of the research, data from 2020 on the development of grape production were not available.

4. Discussion

The Czech Republic is a country with a long tradition of both wine production and consumption. From 1995 to 2016, wine production increased from 459 (in 1000 hL) to 631 (in 1000 hL). Production grew at an average annual rate of 4.57% per year. In contrast, wine consumption in the Czech Republic increased from 1995, when its value was 63.7 (1000 hL), to 192 (1000 hL). Wine production in the Czech Republic is focused more on white wine. The share of white grape varieties in young vineyards in the Czech Republic in 2015–2018 was 92% of the total area of vineyards. Veltlínské zelené, Pálava, and Ryzlink rýnský are the most common varieties in the Czech Republic [36].

The development of the value of grape production is highly dependent on soil quality, climate, temperature, and total precipitation. Temperature (or the intensity of sunlight) is a particularly sensitive variable, as it has a direct effect on the length of the plant’s growing season and the yield of the grapes [12,37]. This conclusion is in general agreement with Van Leeuwen and Darriet [38], who argue that agriculture is highly dependent on climatic conditions during the growing season of crops, which also significantly affects world crop production. These finding were confirmed by Koch and Oehl [39] directly using the example of vines [40]. Current, confirmed, and ongoing climate change mainly affects the volume of production, the sugar content, and, thus, the quality of the final-product wine. Consequently, there is a need to identify ways by which to predict the effects of climate change on crop-growth conditions (in our case, vines) in order to reduce the potential risks posed by climate change in the future, such as the decline in the value of crop production [41,42]. Some authors have already pointed out that, in the future, due to climate change, it will probably be necessary to establish new wine-growing areas in new localities, where vine cultivation was not possible before [6]. The authors also add that this can lead to other risks, such as negative impacts on ecosystems or the loss of biodiversity in the given localities [1]. In addition, the increased use of water for irrigation due to warming already appears to be problematic in a number of areas [43].

Given the ongoing climate change, global warming, and changes in total precipitation, it is necessary to analyze the impact of these climate changes on the production of grapes. This article has analyzed the effect of changes in average temperatures and changes in total precipitation on the production of grapes in the Czech Republic. Secondary data drawn from official sources were used for the analysis, namely, data on the development of the value of vine production and meteorological data on the development of total precipitation and average temperatures. The influence of temperatures and precipitation was investigated during the growing season of the vine.

The multiple regression model did not reveal a common dependence of the total precipitation and average temperatures on the development of the value of vine production, mainly due to the statistical insignificance of the effect of average temperatures on the value of vine production. However, the effect of the change in total precipitation on the value of grape production was confirmed. The results can be interpreted as meaning that if, at the current values and average development of total precipitation, precipitation increases during the growing season of grapes by 1 mm, then the annual production of grapes will decrease on average by 896,976 CZK, and if the average annual temperature in the growing months increases by 1 °C, then the value of grape production will increase on average by 77,293,737 CZK. The increase in grapes’ sugar content in the Czech Republic was also confirmed during the period under review, which will need to be taken into account in further production [40].

The theoretical benefit of the article is the presentation of a summary regarding the effect of average temperatures and total precipitation on the Czech Republic; the practical benefit is to confirm the extent of this effect with its possible use in the form of a prediction of the future development of grape production depending on further climate change.