Authors:
D. Szám Department of Regional Water Management, Faculty of Water Sciences, Ludovika University of Public Service, 6500 Baja, Hungary
National Laboratory for Water Science and Water Security, Faculty of Water Sciences, Ludovika University of Public Service, 6500 Baja, Hungary

Search for other papers by D. Szám in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0009-0006-4299-8712
,
Zs. Hetesi Department of Water and Environmental Security, Faculty of Water Sciences, Ludovika University of Public Service, 6500 Baja, Hungary
Faculty of Natural Sciences, Institute of Mathematics and Informatics, University of Pécs, 7624 Pécs, Hungary

Search for other papers by Zs. Hetesi in
Current site
Google Scholar
PubMed
Close
,
A. P. Takács Department of Plant Protection, Institute of Plant Protection, Hungarian University of Agriculture and Life Sciences, Georgikon Campus, 8360 Keszthely, Hungary

Search for other papers by A. P. Takács in
Current site
Google Scholar
PubMed
Close
,
G. Keve Department of Regional Water Management, Faculty of Water Sciences, Ludovika University of Public Service, 6500 Baja, Hungary

Search for other papers by G. Keve in
Current site
Google Scholar
PubMed
Close
,
P. Balling Tokaj Wine Region's Research Institute for Viticulture and Oenology, 3915 Tarcal, Hungary

Search for other papers by P. Balling in
Current site
Google Scholar
PubMed
Close
, and
Á. Fekete Department of Hydraulic Engineering, Faculty of Water Sciences, Ludovika University of Public Service, 6500 Baja, Hungary

Search for other papers by Á. Fekete in
Current site
Google Scholar
PubMed
Close
Open access

Abstract

This study aims to predict drought periods affecting the Tokaj-Hegyalja wine region and the application of this in crop protection. The Tokaj-Hegyalja wine region is the only closed wine region in Hungary with a specific mesoclimate and a corresponding wine grape variety composition, in which climate change strongly threatens cultivation. The probability that a randomly selected day in the vegetation period will fall into a drought period in the future was estimated using the daily precipitation amount and daily maximum temperature data from the Hungarian Meteorological Service for the period 2002–2020. The Markov model, a relatively new mathematical method for the statistical investigation of weather phenomena, was used for this. Markov chains can, therefore, be a valuable tool for organizing integrated pest management. This can be used to plan irrigation, control fungal pathogens infecting the vines, and plan the success of a given vintage.

Abstract

This study aims to predict drought periods affecting the Tokaj-Hegyalja wine region and the application of this in crop protection. The Tokaj-Hegyalja wine region is the only closed wine region in Hungary with a specific mesoclimate and a corresponding wine grape variety composition, in which climate change strongly threatens cultivation. The probability that a randomly selected day in the vegetation period will fall into a drought period in the future was estimated using the daily precipitation amount and daily maximum temperature data from the Hungarian Meteorological Service for the period 2002–2020. The Markov model, a relatively new mathematical method for the statistical investigation of weather phenomena, was used for this. Markov chains can, therefore, be a valuable tool for organizing integrated pest management. This can be used to plan irrigation, control fungal pathogens infecting the vines, and plan the success of a given vintage.

Introduction and review of literature

Tokaj-Hegyalja has been Hungary's only closed wine region since 1737. As a result of its unique natural features and centuries-old grape and wine culture, it was added to the World Heritage List in 2002 (Hungarian Central Statistical Office, 2016; Szepesi et al., 2017). The authorized wine grape varieties have excellent noble rotting potential (Dékány et al., 2010). Only two varieties (Vitis vinifera cv. Furmint, Vitis vinifera cv. Hárslevelű) are used in 70% of the area, which are relatively drought tolerant due to their efficient stomatal control. However, more frequent droughts from climate change may expose vineyards, especially young vines with still shallow roots, to increased drought stress, thus endangering wine production and quality (Fraga et al., 2012). Viticulture research is essential in assessing possible climate scenarios, estimating the expected changes, and preparing the sector for more sustainable production (Mozell and Thach, 2014). Drought can lead to, among other things, stagnation of protein synthesis, increased soluble protein content, changes in phenolic composition (Bene et al., 2023), as well as affecting grape berry size (Costa et al., 2020; Pirata et al., 2018). Drought can also have a significant effect on grape yields, as one study concluded by showing a statistical relationship between the Standardized Precipitation Index and yields per hectare (Tokhi Arab et al., 2022).

However, the climate and drought adaptation options for vines are limited. The possibilities for variety replacement for adaptation are hampered legally and economically by the long turnaround time (3–5 years) of newly planted vineyards. Solutions that mitigate water scarcity are thus mainly limited to agrotechnology (drought tolerant rootstocks, adoption of intelligent tillage and nutrient management, more intensive summer cover, and shade cultivation) (Keller, 2010; Mozell and Thach, 2014). However, a satisfactory long-term solution to alleviating water scarcity due to climate change can only be provided by easing irrigation and its legal conditions (Tomaz et al., 2015). This is also stated in the new Hungarian domestic law on irrigation management (Act CXIII of 2019). In addition to quantity, timing is also essential when planning irrigation: vines require significant water during the phenological phase, from flowering to veraison (ripening). During this period, both an extreme absence of rainfall and an abundance of rain are undesirable. An abundance of rainfall can open the way to fungal infections (powdery mildew, Peronospora on grape vines, and botrytis grey rot) that cause the most severe economic damage to vines (economic damage to grape production).

In addition to the water requirements of the vines, the production of conidia by Botrytis cinerea, the fungus that plays the most crucial role in noble rot, may also require adaptation to the September-October droughts. While Botrytis cinerea infection before the fruit set can cause grey rot, which is economically damaging, it can also cause noble rot on grapes that are already ripe and can be financially beneficial under suitable mesoclimatic conditions. Botrytis cinerea fungi's conidial production depends on the relative humidity: it is maximum at high values of around 85% (Mesterházy et al., 2014; Ciliberti et al., 2016). One study found that the choice of grape variety had no significant effect on conidia production. In contrast, there is a significant effect of weather (vintage, month, day), especially the frequency and length of dry, droughty periods. To this end, adoption experiments have been carried out with artificial humidification.

There are typically one or two years out of ten in the Tokaj hills when wineries rate the botrytized dessert wine (called aszú wine) yield as good. More typical is the average yield, which is not threatened by prolonged droughts or rainy periods. Since the quality of the crop is significantly affected by weather conditions, much agrotechnical research aims to find ways to reduce weather exposure for better yields. To this end, adoption experiments have been carried out with artificial humidification.

Part of effective production is the use of longer-term climate forecasts as well as short-term weather forecasting models. It is crucial for farmers to know whether a drought is going to continue. Forecasting this using probabilistic estimation is possible on a statistical basis and offers economic benefits. A relatively simple statistical model using Markov chains can help us with this. The field of weather forecasting employs a multitude of sophisticated mathematical tools that consume a considerable amount of computing power and data to process. The Markov process, however, is unable to compete with these methods. Nevertheless, in the simplified case of calculating only the probability of the next day being droughty, its simplicity and lower computational requirements make it a worthy consideration (see e.g. Held – Sabanés, 2020).

The main objective of our manuscript is the statistical analysis of drought periods, including the analysis with Markov chains. In viticulture, the monitoring of drought periods is important in many ways. A study has developed a drought index specific to grapes, highlighting that the investigation of process-based meteorological evaluation indicators is particularly interesting in precisely understanding the spatiotemporal characteristics of grape drought processes (Huo et al., 2022). Summing up the above, drought monitoring is important in the selection of varieties, in the application of agrotechnological solutions that treat water shortages, and in the planning of any necessary artificial water replacement. It may point to the need for artificial humidification experiments. In addition to B. cinerea, which causes noble rot and gray rot, many other fungal infections damage grapes, and drought monitoring is an important condition for successful forecasting. If we know the temperature and precipitation conditions through drought monitoring, according to our hypothesis, it can bring us closer to predicting the infection of plant pathogenic and sometimes toxinogenic fungi. Plant pathogenic fungi's development, reproduction, and toxin production have a well-defined temperature, humidity range, and optimum (Garcia-Cela et al., 2018; Medina et al., 2017; Wu et al., 2011). The monitoring of these abiotic factors of infections in local production in an IoT (Internet of Things) system, with measurement evaluation software, can help plant protection decision-making in the future (Balaceanu et al., 2021; Liopa-Tsakalidi et al., 2021). Markov chains have the advantage of being able to model complex systems relatively simply, as they do not require detailed information about the system's history. Their application requires only knowledge of the system's current state (Aaron, 2023; Meyn and Tweedie, 2012). Markov chains are functional in the statistical analysis of various weather events. They have been used to predict the probability of drought (Fekete, 2022a; Freidoni, 2015), forest fires (Martell, 1999), land-use changes (Zhang et al., 2011), rainfall sums (Fekete and Keve, 2020; Ibeje et al., 2018; Tettey et al., 2017), and flash floods (Fekete, 2022b) and general weather chains (Gates and Tong, 1976; Mississippi, 2022). According to our literature review, only two studies (Fekete, 2022a; Freidoni, 2015) have so far attempted to analyze droughts in agriculture with Markov chains, which gives the novelty of this paper.

Any statistical tool capable of predicting the probability of different extremes could be an important area of research. Thus, for Markov processes, calculating a limiting distribution, which is equivalent to the long-term probability of occurrence of the elements of the process, is of particular importance. Markov chains are useful in two ways. On the one hand, the forecasts are available immediately after the observation is completed, since only local weather information is used as a forecaster. On the other hand, they require relatively few calculations after processing the climatological data.

However, there are limitations to the use of Markov chains. Some methodological limits are shown by applying the Markov chain in weather forecasting, especially drought. Among these is the reliance on simplified assumptions, where the Markov chains are assumed to forecast the current state of a system alone, which is not a true reflection of probabilities considering the sequence of preceding events. The result can be the oversimplification of the multifaceted nature of weather. However, this can be modified by fine-tuning the model to account for cross-correlations between weather variables and soil monitoring. Second, Markov chains have long-term drawbacks. Markov chains are more efficient for short-term forecasts, i.e., their applicability is reduced for long-term forecasts, especially when the weather system changes, e.g., in the wake of climate change. It should be noted that in our case, the Markov model is sufficient for the short-term prediction of drought days. The distant parts of the data series do not significantly affect the following day's drought.

Material and methods

The study area (Fig. 1) was the Tokaj-Hegyalja wine region. It is bordered to the northwest by the Hernád River and to the southeast by the Tisza and its tributary, the Bodrog. This triangular region of more than 88,000 ha covers the administrative territory of 27 municipalities.

Fig. 1.
Fig. 1.

The Tokaj hegyalja wine region with the municipalities of Tarcal and Sátoraljaújhely, where the meteorological stations are located

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

In the Tokaj-Hegyalja wine region, only two continuously maintained automatic hydro-meteorological measuring stations provide data suitable for scientific research. One is in the village of Tarcal, bordered by HungaroMet, measuring station No. 62612. Another is in Sátoraljaújhely, bordered by HungaroMet measuring station No. 61709. The daily precipitation sum and maximum temperature data from these stations (HungaroMet, 2023) were used from 1 March to 31 October from 2002 to 2020.

More available measuring stations outside the wine region are unlikely to provide additional data for analysis. From the daily maximum temperature and precipitation data for Tarcal and Sátoraljaújhely, we found that the patterns are not normally distributed using Shapiro-Wilk tests. Since the theoretical distribution of the samples is unknown, they were compared using two-sample Kolmogorov-Smirnov tests for each year between 2002 and 2020. This statistical test compares the continuous distribution of two samples. This test does not impose any conditions on the distributions of the samples. The null hypothesis of the test is that the two samples come from the same distributions.

We used a significance level lower than the usual α = 0.05 because the two-sample Kolmogorov‒Smirnov tests are more robust, weaker statistical tests (Marozzi, 2009, 2013). All statistical tests were generated using Wolfram Mathematica 9 software.

According to Article 2(1) of Hungarian Act CLXVIII of 2011, drought is defined as a natural event during which the total amount of precipitation falling on the place of risk in the growing season of the crop concerned is less than 10 mm for 30 consecutive days, or the rainfall total that falls is less than 25 mm. The daily maximum temperature exceeds 31 Celsius degrees on at least 15 days. In this study, drought periods were established based on two criteria that apply from 2017. To establish the existence of a risk of damage, one of these two conditions must be fulfilled, which must be supported by data from the HungaroMet.

A Markov chain is a stochastic process Xt, such as the probability that Xt+1 takes a value j at time t+1 depends on the past only through its most recent value Xt at time t:
P(Xt+1=j|X0=x0,X1,,Xt1=xt1)=P(Xt+1=j|Xt=i)=pij
for any of the i,jS and t T (Paulo and Perreira, 2007). The probability of transition from point i to point j is pij. The transition probability can be expressed as a matrix as follows
Pij=[p11p12p1np21p22p2n......pn1pn2pnn].
where 0pij1 and jpij=1;i=1,2,n. In long-term processes, the transition probabilities of the Markov chain states are independent of the initial state. Therefore, the limiting distribution could be calculated by successive multiplications of the transition matrix itself (Dobrow, 2016; Ross, 2010). The understanding of Markov processes, the transition probability matrix and the limiting distribution can be illustrated with an example. For example, a summer day is more likely to be clear the next day. Let the clear sky be state 1, and the overcast sky be state 2. For example, if we construct a one-step transition matrix from the data of a summer, it is:
Pij=(0.80.20.60.4)
This means: that if the sun was shining on one day (first row), the probability of sunshine on the next day is 0.8 and the probability of overcast skies is 0.2. If it was overcast on a given day (second row), the probability of sunshine on the next day is 0.6 and the probability of overcast skies is 0.4. If we choose a day with sunshine as a starting point, we multiply this matrix from the left by the state vector P = (1,0), and the resulting vector for the next day shows the probability of the two states: P+1 = (0.8,0.2). If the transition matrix is raised to the second power, then multiplying by the same row vector gives the probabilities two days from now, and so on. In the long run, regardless of the initial weather, the probability that it will be sunny on any given day is 2/3. This is obtained by raising the transition matrix to high power and transforming it into two identical row vectors with identical elements. This is the limiting distribution, which is otherwise equal to the frequency that can be calculated from all the data. The Markov model was tested as follows. The limiting likelihoods were established from the HungaroMet data series. We distinguished between drought and non-drought days. The transition probability matrix (H) for a given period of the year under study was generated as follows. Based on the HungaroMet measurements, we examined the value that follows a value for a given time interval. If a humidity class i ∈ {0; 1} was measured in a given measurement and j ∈ {0; 1} in the subsequent measurement, the value of the element hij of the transition probability matrix increases by one. We defined a drought day as state 0 and a non-drought day as state 1. The transition probability matrix for a given year is then:
H=(h00h01h10h11)

Using the formula (Eq. (2)), we calculated the transition probability for each year and then determined the limiting distribution.

Results and discussion

By law, we have defined the drought periods between 2002 and 2020 (Table 1, Fig. 2). There was at least one drought period in fifteen of the examined years. In seven cases, two drought periods occurred within a year. 2020 had the most drought days within two drought periods (118 days), followed by 2015, with two drought periods (99 days total). These two years mostly favored botrytis-free, full-bodied, dry wines, and aszú became very scarce in the entire wine region. The appearance of botrytis was most favorable in those years when the drought ended in the first half of September at the latest. After August, the drought-free, moderately rainy autumn weather was already favorable for the appearance and spread of botrytis, which causes noble rot. In none of the best vintages (2003, 2006, 2007, 2013, 2017, 2019) did a drought occur from the second half of September.

Table 1.

Drought periods 2002−2020 as determined under Act CLXVIII of 2011, based on data from the HungaroMet measuring station in Tarcal

Year2002200320042005200620072008200920102011201220132014201520162017201820192020
Start and end of periods13/8

14/9
1/3

22/4
1/3

14/4
28/6

28/7
24/3

4/5
14/3

28/5
6/9

6/10
1/3

5/4
10/4

11/5
1/4

1/5
8/3

10/5
2/3

5/4
2/4

15/5
1/3

21/4
3/3

23/5
31/8

28/10
5/8

3/9
13/8

14/9
12/7

24/8
30/6

5/9
19/7

25/8
20/8

24/9

The numbers in the cells are dates in format d/m.

Source: Authors.

Fig. 2.
Fig. 2.

Length in days of drought periods 2002−2020, from 1 January to the last day of the year, based on data from the HungaroMet measuring station in Tarcal. First drought period within a year is noted by gray, second by brown color. Source: Authors

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

Out of the 19 years examined, there were only four years (2004, 2008, 2010, 2014) when there was no drought. The lack of drought in these years was caused by the precipitation and its even distribution within the vegetation period. In these years, mainly due to gray rot and other pathogenic epidemics that prefer cool, moist environments, the aszú wine harvest was small.

The frequency of drought days (2002–2020) gave a frequency distribution plot with two local maxima. A frequency maximum can be observed in the second half of March and the first half of April. This is mainly caused by the lack of precipitation and the uneven distribution of daily precipitation amounts. The daily maximum temperature (with the exception of a few summer days in May) played a smaller role in this. It was also interesting to observe the very early, late February and early March drought days, which are mentioned by several sources as damaging weather phenomena. Dry springs increase the likelihood of extremely hot and dry summers due to feedback from soil moisture and increased drought impacts (Haslinger and Blöschl, 2017; Mueller and Seneviratne, 2012).

Another lower but longer frequency maximum can be observed in August's second half and September's first half. In addition to the lack of precipitation and the unfavorable distribution of daily precipitation amounts, heat also plays a decisive role in this (Fig. 3). In the 19 years of the study, the late summer drought period, which often extended into September, was not uncommon, with days with summer temperatures in addition to the lack of precipitation. According to several Mediterranean forecasts, summer droughts are expected to become drier and last longer, extending beyond the summer (i.e., toward autumn, spring, or both) (Gibelin and Déqué, 2003; Giorgi and Lionello, 2008).

Fig. 3.
Fig. 3.

Frequency (%) of drought days between 2002 and 2020, from 1 January to the last day of the year, based on data from the HungaroMet measuring station in Tarcal. Source: Authors

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

The frequency of droughts increased in the second period, 2012–2020. Comparing the drought frequencies in the 2003–2011 and 2012–2020 periods using a two-sample Kolmogorov‒Smirnov test, we found that the distributions were significantly different. This means that the temporal pattern of droughts has changed significantly. However, this result should be treated with caveats, given that precipitation (both in terms of quantity and distribution) shows relevant variability from year to year.

For 15 of the 19 years studied, the drought periods were caused by a lack of rainfall and an unequal distribution of daily rainfall amounts. In the 2012–2020 period, total rainfall decreased by 8.14% compared to the 2003–2011 period. However, the change in rainfall distribution by volume is even higher.

The average monthly rainfall decreased in March, April, May, June, and August, while it increased slightly in July, September, and October, but none of the changes was significant (Fig. 4). Comparing the results with national data from HungaroMet, it can be seen that the average rainfall totals for the months of March and April in the period 2002–2020 were below the national averages for the period 1991–2020 (HungaroMet Hungarian Meteorological Service Nonprofit Private Limited Company data, 2021a), indicating the importance of early spring droughts. This is significant because Tarcal is typically rainier than the country average. The average monthly rainfall in May, which is critical for cereal crops, decreased the most (−29.87%) in the second period. In August and September, the average rainfall totals for the period 2002–2020 were below the country average for the period 1991–2020 (HungaroMet, 2021b).

Fig. 4.
Fig. 4.

Average monthly rainfall between 2002 and 2020 from March to October, based on data from the HungaroMet measuring station in Tarcal. The horizontal axis shows the month under study and the vertical axis the average of the monthly precipitation totals for the periods under study. Source: Authors

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

The number of dry days, when the daily rainfall was more than 1 mm, increased by 3.51%. There was no difference in the number of days with more than 20 mm of daily precipitation. However, the frequency of rainfall-free periods varied. In the second period (2012–2020), more long periods without rainfall occurred.

In four years (2012, 2015, 2017, 2018), the daily maximum temperature was the dominant factor for drought occurrence. When it was fulfilled, the daily maximum exceeded 31 degrees Celsius for at least 15 days within 30 days. Notable among these is 2015, which hit several temperature records. There were 33 days with daily maximum temperatures above 31 degrees Celsius during the 30-day period starting with that date. But it is noteworthy that all four years were in the second half of the 19-year period studied. Looking at the period 2003–2020, it can be concluded that the impact of daily maximum temperature on drought occurrence may increase in the future (Fig. 5). In the years 2003–2011, there were 125 days with daily maximum temperatures above 31 °C. In contrast, there were 204 such days between 2012 and 2020. Furthermore, a two-sample Kolmogorov‒Smirnov test revealed a significant difference in the distribution of daily maximum temperatures of at least 31 degrees Celsius between 2003–2011 and 2012–2020 (Ds=0.231;Ks=1.44;ps=0.031<0.05 where Ds is the largest distance between the two distributions and Ks is the critical value).

Fig. 5.
Fig. 5.

Frequency distribution of daily maximum temperatures over 31 Celsius degrees for the periods 2003–2011 and 2012–2020 at decimal degree resolution. The horizontal axis shows the temperature in tenths of a degree Celsius and the vertical axis shows the frequency of occurrence of the temperature data. Source: Authors

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

The 19 years studied are not long enough to show the weather effects of global climate change. The results should, therefore, be reviewed for longer time horizons. Their spatial validity is limited by the fact that the distribution of the precipitation data measured at the two stations (Tarcal and Sátoraljaújhely) is significantly different. As already pointed out above, the rainfall, not the daily maximum temperature, is the determining factor in the occurrence of droughts in the sense of the Hungarian legislation.

After defining the drought periods, the transition probability matrices for each year (Table 2) were determined using formula (3). Then, the limiting probabilities P1 and P2 were calculated (Fig. 6), but one might note that the same is available directly from the initial frequencies themselves as well. The probability of limiting distribution of a given year only estimates how likely a day is to be a drought within a year. We can see that 2020, 2015, and 2006 had the highest probability of a randomly chosen day of the growing season falling in a drought period.

Table 2.

Transition probability matrices (P) for the heatwave for the period 2002−2020 were calculated from the HungaroMet measurements

Vegetation periodPVegetation periodP
2002(2101132)2012(1741267)
2003(1901152)2013(1662274)
2004(244000)2014(244000)
2005(1981144)2015(1432297)
2006(1522288)2016(1791163)
2007(1702270)2017(2081134)
2008(244000)2018(1602280)
2009(1671175)2019(1911151)
2010(244000)2020(12422116)
2011(2121130)

Source: Authors.

Fig. 6.
Fig. 6.

P1 and P2 are limiting distributions for the drought test for the period 2002−2020, calculated from the HungaroMet measurements. Source: Authors

Citation: Progress in Agricultural Engineering Sciences 20, 1; 10.1556/446.2024.00116

Over a 19-year period, we find that the average P2 limiting probability is 20.8%. This means that, on average, there is a 20.8% chance of any day during the vegetation season (from 1 March to 31 October) falling within a drought period.

We did not have prior knowledge of the mean and standard deviation, so we used Shapiro‒Wilk tests to examine the normality of the P1 and P2 values. The result was that neither of the P1 and P2 values can be described by a normal distribution (P1: W(19) = 0.87, P = 0.017; P2: W(19) = 0.87, P = 0.017). This implies that Markov chains have limited applicability beyond a year for predicting droughts.

We have found that the drought periods identified at Tarcal are only conditionally valid for the Tokaj-Hegyalja wine region. The Kolmogorov‒Smirnov tests did not show a significant difference in the distribution in either year when comparing the daily maximum temperature data in Tarcal and in Sátoraljaújhely (Table 3).

Table 3.

Comparison of daily maximum temperature and precipitation sum data from Tarcal and Sátoraljaújhely using Kolmogorov-Smirnov tests (2002–2020). Statistical significance levels: *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. Significantly different distributions are indicated in bold type in the table

YearPrecipitation P [mm]Daily maximum temperature Tmax [°C]
nDsKsnDsKs
20022450.1825***0.17612450.0048***0.1761
20032450.3488***0.17612450.0049***0.1761
20042450.2192***0.17612450.0070***0.1761
20052450.1500**0.14712450.0037***0.1761
20062450.2820***0.17612450.0026***0.1761
20072450.1848***0.17612450.0047***0.1761
20082450.1787***0.17612450.0047***0.1761
20092450.1651***0.17612450.0050***0.1761
20102450.2720***0.17612450.0052***0.1761
20112450.2999***0.17612450.0055***0.1761
20122450.1233*0.12272450.0041***0.1761
20132450.4303***0.17612450.0044***0.1761
20142450.1320***0.17612450.0054***0.1761
20152450.4128***0.17612450.0026***0.1761
20162450.1684**0.14712450.0030***0.1761
20172450.2133***0.17612450.0026***0.1761
20182450.3102***0.17612450.0028***0.1761
20192450.1806***0.17612450.0050***0.1761
20202450.2144***0.17612450.0049***0.1761

Source: Authors.

In contrast, the distribution of daily precipitation data differed significantly between the two weather stations in the 19 years studied (Table 3). This is not a surprising result, given precipitation's rapid temporal and spatial variability. It shows that our study and its subsequent results have limited applicability to the Tokaj-Mountain Valley wine regions.

Conclusions

In summary, for the twenty years of weather studies, it can be shown that the drought stress effect typically occurs in two periods: from March to May and from July to September. In the former period, the probability of occurrence of early spring droughts increases significantly with the years. In the second period, droughts are earlier and their frequency increases in August. This shift may result in earlier ripening and subsequent earlier harvest. However, further studies on artificial humidification in autumn are worthwhile without an earlier harvest. The increase in the frequency and severity of droughts in August points to the need for artificial watering this month. The more frequent occurrence of droughts indicates the importance of selecting drought-tolerant grape varieties and breeding for drought stress resistance.

The lack of rainfall was primarily responsible for the droughts. In only four of the twenty years studied, high daily maximum temperatures were responsible for drought.

The calculations show that the Markov model is suited to studying droughts. The results can be used to determine the probability that a given drought day will be followed by another drought day. The Markov model's validity over time is that it is not suitable for longer-term forecasts beyond the annual time scale for a 20-year data set. Longer time series may be required for forecasting beyond one year.

An examination of the spatial validity of the model suggests that Validity checks of Markov chains are necessary when examining droughts over an area larger than a city or district. We have seen that daily precipitation is of great importance in the occurrence of droughts. Precipitation shows a different pattern for two Hungarian automatic gauging stations, typically located at least 20–30 km away in Hungary. Exceptions are more prominent cities, with several stations in a town. Exceptions are also riverside settlements, where the network of hydro-meteorological stations is denser due to flood protection.

Using Markov-type prediction in drought evolution seems to be a good choice because the onset and persistence of drought are crucially determined by processes not in the data series of the distant past, but in the preceding step, the Markov chain is memory-less. It should be noted that the limiting distribution will, of course, reflect the annual frequency of drought days, but the one-step transition matrix is a good predictor for farmers. Suppose a forecasting system produces a one-step transition matrix online for a given day of the year. In that case, a better estimate of the status of the next day's drought, for example, can be obtained from the status of that day than the probability determined from the frequency. By increasing the transition matrix to the second, third, etc. power, an estimate of the probability for two or three days in advance can be obtained. The further forward in time we go, the more we move towards a probability derived from frequency and the advantage of the Markov model disappears.

Further use of Markov chains and fine-tuning of the model (more accurate spatial and temporal validation, analysis of specific phenological phases and fungal infection periods) can help to organize integrated pest management. This includes precision irrigation and determining the right time to control fungal pathogens that infect vines. Accurate forecasting is the basis for environmentally friendly crop protection, which can significantly reduce the use of pesticides.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements

The research presented in the article was carried out within the framework of the Hungarian Széchenyi Plan Plus program with the support of the RRF 2.3.1 21 2022 00008 project.

References

  • Aaron, K. (2023). Research on the application of Markov chain. International Conference on Mathematical Physics and Computational Simulation, Oxford, United Kingdom, Aug 12‒18 2023, https://doi.org/10.54254/2753-8818/11/20230395.

    • Search Google Scholar
    • Export Citation
  • Balaceanu, C.M., Dragulinescu, A.M.C., Orza, O., and Bosoc, S. (2021). Monitoring the vineyard health using internet of things sensors in smart agriculture – a technical report. Conference: Air and Water – Components of the Environment 2021 Conference Proceedings, 1: 131140, https://www.doi.org/10.24193/AWC2021_12.

    • Search Google Scholar
    • Export Citation
  • Bene, Zs., Kiss, I., and Balling, P. (2023). A klímaváltozás borkémiai hatásainak vizsgálata a Tokaji borvidéken [Investigation of the wine chemical effects of climate change in the Tokaj wine region]. In: Bene, Zs. (Ed.), THE Eszencia – Bor és Tudomány: Fejezetek a Lorántffy Intézet oktatóinak tollából. Tokaj-Hegyalja Egyetem.

    • Search Google Scholar
    • Export Citation
  • Ciliberti, N., Fermaud, M., Roudet, J., Languasco, L., and Rossi, V. (2016). Environmental effects on the production of Botrytis cinerea conidia on different media, grape bunch trash, and mature berries. Australian Journal of Grape and Wine Research, 22(2): 262270, https://doi.org/10.1111/ajgw.12217.

    • Search Google Scholar
    • Export Citation
  • Costa, C., Graça, A., Fontes, N., Teixeira, M., Gerós, H., and Santos, J.A. (2020). The interplay between atmospheric conditions and grape berry quality parameters in Portugal. Applied Sciences, 10(14): 122, https://doi.org/10.3390/app10144943.

    • Search Google Scholar
    • Export Citation
  • Dékány, T. and Técsi, Z. (2010). A tokaji borvidék – Világörökség. Corvina Kiadó, Budapest.

  • Dobrow, R.P. (2016). Introduction to stochastic processes. John Wiley and Sons, New York.

  • Fekete, Á. (2022a). A villámárvíz valószínűségének becslése kétváltozós Markov-lánccal [Estimation of the probability of flash flooding using a bivariate Markov chain]. Hidrológiai Közlöny, 102(2): 4149, (in Hungarian).

    • Search Google Scholar
    • Export Citation
  • Fekete, Á. (2022b). Markov chain analysis of the probability of days in the heatwave period. Időjárás − Quarterly Journal of the Hungarian Meteorological Service, 126(3): 375386, https://doi.org/10.28974/idojaras.2022.3.6.

    • Search Google Scholar
    • Export Citation
  • Fekete, Á. and Keve, G. (2020). A csapadékösszegek és az aszályos időszakok vizsgálata Markov-láncokkal [Study of precipitation totals and periods of drought using Markov chains]. Hidrológiai Közlöny, 100(4): 6070, (in Hungarian).

    • Search Google Scholar
    • Export Citation
  • Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., and Santos, J.A. (2012). An overview of climate change impacts on European viticulture. Food Energy Security, 1(2): 94110, https://doi.org/10.1002/fes3.14.

    • Search Google Scholar
    • Export Citation
  • Freidooni, F., Ataei, H., and Shahriars, F. (2015). Estimating the occurrence probability of heat wave periods using the Markov chain model. Journal of Sustainability Development, 8(2): 2645, https://doi.org/10.5539/jsd.v8n2p26.

    • Search Google Scholar
    • Export Citation
  • Garcia-Cela, E., Verheecke-Vaessen, C., Magan, N., and Medina, A. (2018). The “-omics” contributions to the understanding of mycotoxin production under diverse environmental conditions. Current Opinion in Food Science, 23: 97104, https://doi.org/10.1016/j.cofs.2018.08.005.

    • Search Google Scholar
    • Export Citation
  • Gates, P. and Tong, H. (1976). On Markov chain modeling to some weather data. Journal of Applied Meteorology, 15(11): 11451151, https://doi.org/10.1175/1520-0450(1976)015<1145:OMCMTS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gibelin, A.L. and Déqué, M. (2003). Anthropogenic climate change over the Mediterranean region simulated by a global variable resolution model. Climate Dynamics, 20(4): 327339, https://doi.org/10.1007/s00382-002-0277-1.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F. and Lionello, P. (2008). Climate change projections for the Mediterranean region. Global and Planetary Change, 63(2‒3): 90104, https://doi.org/10.1016/j.gloplacha.2007.09.005.

    • Search Google Scholar
    • Export Citation
  • Haslinger, K. and Blöschl, G. (2017). Space-time patterns of meteorological drought events in the European greater alpine region over the past 210 years. Water Resources Research, 53(11): 98079823, https://doi.org/10.1002/2017wr020797.

    • Search Google Scholar
    • Export Citation
  • Held, L. and Sabanés Bové, D. (2020). Markov models for time series analysis. In: Likelihood and Bayesian inference. Statistics for biology and health. Springer, Berlin, Heidelberg, https://doi.org/10.1007/978-3-662-60792-3_10.

    • Search Google Scholar
    • Export Citation
  • Hungarian Central Statistical Office (2016). The World Heritage Tokaj wine region, Available at: https://www.ksh.hu/docs/hun/xftp/idoszaki/pdf/tokaji_borvid.pdf (Accessed 13 October 2023).

    • Search Google Scholar
    • Export Citation
  • HungaroMet (2023). Meteorológiai Adattár (Meteorological Database). Available at: https://odp.met.hu/

  • HungaroMet (Hungarian Meteorological Service Nonprofit Private Limited Company) data (2021a). Available at: https://www.met.hu/eghajlat/eghajlatvaltozas/megfigyelt_hazai_valtozasok/homerseklet_es_csapadektrendek/csapadekosszegek/.

  • HungaroMet (Hungarian Meteorological Service Nonprofit Private Limited Company) data (2021b). Available at: https://www.met.hu/eghajlat/magyarorszag_eghajlata/altalanos_eghajlati_jellemzes/csapadek/.

  • Huo, Z., Mao, H., Yang, J., and Wang, P. (2022). Process-based evaluation indicators of grape drought and risk characteristics in the Bohai Rim Region, China. Theoretical and Applied Climatology, 150(1): 126, https://doi.org/10.21203/rs.3.rs-1724263/v1.

    • Search Google Scholar
    • Export Citation
  • Ibeje, A.O., Osuagwu, J., and Onosakponome, R. (2018). A Markov model for prediction of annual rainfall. Journal of Engineering and Applied Science, 3(11): 15.

    • Search Google Scholar
    • Export Citation
  • Keller, M. (2010). Managing grapevines to optimize fruit development in a challenging environment: a climate change primer for viticulturists. Australian Journal of Grape and Wine Research, 16(s1): 5669, https://doi.org/10.1111/j.1755-0238.2009.00077.x.

    • Search Google Scholar
    • Export Citation
  • Liopa-Tsakalidi, A., Thomopoulos, V., Pantelis, B., and Kavga, A. (2021). A NB-IoT based platform for smart irrigation in vineyard. Conference: 10th International Conference on Modern Circuits and Systems Technologies (MOCAST). https://www.doi.org/10.1109/MOCAST52088.2021.9493381.

    • Search Google Scholar
    • Export Citation
  • Marozzi, M. (2009). Some notes ont the location-scale Cucconi test. Journal of Nonparametric Statistics, 21(5): 629647, https://doi.org/10.1080/10485250902952435.

    • Search Google Scholar
    • Export Citation
  • Marozzi, M. (2013). Nonparametric simultaneous tests for location and scale testing: a comparison of several methods. Communications in Statistics – Simulation and Computation, 42(6): 12981317, https://doi.org/10.1080/03610918.2012.665546.

    • Search Google Scholar
    • Export Citation
  • Martell, D.L. (1999). A Markov chain model of day-to-day changes in the Canadian forest fire weather index. International Journal of Wildland Fire, 9(4): 265273, https://doi.org/10.1071/WF00020.

    • Search Google Scholar
    • Export Citation
  • Medina, A., Gilbert, M.K., Mack, B.M., O’Brian, G.R., Rodríguez, A., Bhatnagar, D., Payne, G., and Magan, N. (2017). Interactions between water activity and temperature on the Aspergillus flavus transcriptome and aflatoxin B1 production. International Journal of Food Microbiology, 256: 3644.

    • Search Google Scholar
    • Export Citation
  • Mesterházy, I., Mészáros, R., and Pongrácz, R. (2014). The effects of climate change on grape production in Hungary. Időjárás − Quarterly Journal of the Hungarian Meteorological Service, 118(3): 193206.

    • Search Google Scholar
    • Export Citation
  • Meyn, S.P. and Tweedie, R.L. (2012). Markov chains and stochastic stability. Springer Science & Business Media, London.

  • Mississippi, V. (2022). Markov chains and applications. Selecciones Matemáticas, 9(1): 5378, https://doi.org/10.17268/sel.mat.2022.01.05.

    • Search Google Scholar
    • Export Citation
  • Mozell, M.R. and Thach, L. (2014). The impact of climate change on the global wine industry: challenges and solutions. Wine Economics and Policy, 3(2): 8189, https://doi.org/10.1016/j.wep.2014.08.001.

    • Search Google Scholar
    • Export Citation
  • Mueller, B. and Seneviratne, S.I. (2012). Hot days induced by precipitation deficits at the global scale. Proceedings of the National Academy of Sciences, 109(31): 1239812403, https://doi.org/10.1073/pnas.1204330109.

    • Search Google Scholar
    • Export Citation
  • Paulo, A.A. and Perreira, L.S. (2007). Prediction of SPI drought class transitions using Markov chains. Water Resources Management, 21(10): 18131827, https://doi.org/10.1007/s11269-006-9129-9.

    • Search Google Scholar
    • Export Citation
  • Pirata, M.S. (2018). Estudo Do Stress Hídrico Da Vinha-Castas Aragonês e Trincadeira [Study of Water Stress in Aragonês and Trincadeira Vineyards], Phd thesis. Universidade de Évora, (in Portuguese).

    • Search Google Scholar
    • Export Citation
  • Ross, S.M. (2010). Introduction to probability models. Academic Press, New York.

  • Szepesi, J., Harangi, Sz., Ésik, Zs., Novák, T.J., Lukács, R., and Soós, I. (2017). Volcanic geoheritage and geotourism perspectives in Hungary: a case of a UNESCO World heritage site, Tokaj wine region historic cultural landscape, Hungary. Geoheritage, 9(3): 329349, https://doi.org/10.1007/s12371-016-0205-0.

    • Search Google Scholar
    • Export Citation
  • Tettey, M., Oduro, F.T., Adedida, D., and Abaye, D.A. (2017). Markov chain analysis of the rainfall patterns of five geographical locations on the south-eastern coast of Ghana. Earth perspectives − Transdisciplinary Enabled, 4(6): 111, https://doi.org/10.1186/s40322-017-0042-6.

    • Search Google Scholar
    • Export Citation
  • Tokhi Arab, S., Noguchi, R., and Ahamed, T. (2022). Yield loss assessment of grapes using composite drought index derived from landsat OLI and TIRS datasets. Remote Sensing Applications Society and Environment, 26(3): 100727, https://www.doi.org/10.1016/j.rsase.2022.100727.

    • Search Google Scholar
    • Export Citation
  • Tomaz, A., Martinez, J.M.C., and Pacheco, C.A. (2015). Yield and quality responses of “Aragonez” grapevines under deficit irrigation and different soil management practices in a mediterranean climate. Ciencia e Tecnica Vitivinicola, 30(1): 920, https://doi.org/10.1051/ctv/20153001009.

    • Search Google Scholar
    • Export Citation
  • Wu, F., Bhatnagar, D., Bui-Klimke, T., Carbone, I., Hellmich, R., Munkvold, G., Paul, P., Payne, G., and Takle, E. (2011). Climate change impacts on mycotoxin risks in US maize. World Mycotoxin Journal, 4(1): 7993, https://doi.org/10.3920/WMJ2010.1246.

    • Search Google Scholar
    • Export Citation
  • Zhang, R., Tang, C., Ma, S., Yuan, H., Gao, L., and Fan, W. (2011). Using Markov chains to analyze changes in wetland trends in arid Yinchuan Plain, China. Mathematical and Computer Modelling, 54(3–4): 924930.

    • Search Google Scholar
    • Export Citation
  • Aaron, K. (2023). Research on the application of Markov chain. International Conference on Mathematical Physics and Computational Simulation, Oxford, United Kingdom, Aug 12‒18 2023, https://doi.org/10.54254/2753-8818/11/20230395.

    • Search Google Scholar
    • Export Citation
  • Balaceanu, C.M., Dragulinescu, A.M.C., Orza, O., and Bosoc, S. (2021). Monitoring the vineyard health using internet of things sensors in smart agriculture – a technical report. Conference: Air and Water – Components of the Environment 2021 Conference Proceedings, 1: 131140, https://www.doi.org/10.24193/AWC2021_12.

    • Search Google Scholar
    • Export Citation
  • Bene, Zs., Kiss, I., and Balling, P. (2023). A klímaváltozás borkémiai hatásainak vizsgálata a Tokaji borvidéken [Investigation of the wine chemical effects of climate change in the Tokaj wine region]. In: Bene, Zs. (Ed.), THE Eszencia – Bor és Tudomány: Fejezetek a Lorántffy Intézet oktatóinak tollából. Tokaj-Hegyalja Egyetem.

    • Search Google Scholar
    • Export Citation
  • Ciliberti, N., Fermaud, M., Roudet, J., Languasco, L., and Rossi, V. (2016). Environmental effects on the production of Botrytis cinerea conidia on different media, grape bunch trash, and mature berries. Australian Journal of Grape and Wine Research, 22(2): 262270, https://doi.org/10.1111/ajgw.12217.

    • Search Google Scholar
    • Export Citation
  • Costa, C., Graça, A., Fontes, N., Teixeira, M., Gerós, H., and Santos, J.A. (2020). The interplay between atmospheric conditions and grape berry quality parameters in Portugal. Applied Sciences, 10(14): 122, https://doi.org/10.3390/app10144943.

    • Search Google Scholar
    • Export Citation
  • Dékány, T. and Técsi, Z. (2010). A tokaji borvidék – Világörökség. Corvina Kiadó, Budapest.

  • Dobrow, R.P. (2016). Introduction to stochastic processes. John Wiley and Sons, New York.

  • Fekete, Á. (2022a). A villámárvíz valószínűségének becslése kétváltozós Markov-lánccal [Estimation of the probability of flash flooding using a bivariate Markov chain]. Hidrológiai Közlöny, 102(2): 4149, (in Hungarian).

    • Search Google Scholar
    • Export Citation
  • Fekete, Á. (2022b). Markov chain analysis of the probability of days in the heatwave period. Időjárás − Quarterly Journal of the Hungarian Meteorological Service, 126(3): 375386, https://doi.org/10.28974/idojaras.2022.3.6.

    • Search Google Scholar
    • Export Citation
  • Fekete, Á. and Keve, G. (2020). A csapadékösszegek és az aszályos időszakok vizsgálata Markov-láncokkal [Study of precipitation totals and periods of drought using Markov chains]. Hidrológiai Közlöny, 100(4): 6070, (in Hungarian).

    • Search Google Scholar
    • Export Citation
  • Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., and Santos, J.A. (2012). An overview of climate change impacts on European viticulture. Food Energy Security, 1(2): 94110, https://doi.org/10.1002/fes3.14.

    • Search Google Scholar
    • Export Citation
  • Freidooni, F., Ataei, H., and Shahriars, F. (2015). Estimating the occurrence probability of heat wave periods using the Markov chain model. Journal of Sustainability Development, 8(2): 2645, https://doi.org/10.5539/jsd.v8n2p26.

    • Search Google Scholar
    • Export Citation
  • Garcia-Cela, E., Verheecke-Vaessen, C., Magan, N., and Medina, A. (2018). The “-omics” contributions to the understanding of mycotoxin production under diverse environmental conditions. Current Opinion in Food Science, 23: 97104, https://doi.org/10.1016/j.cofs.2018.08.005.

    • Search Google Scholar
    • Export Citation
  • Gates, P. and Tong, H. (1976). On Markov chain modeling to some weather data. Journal of Applied Meteorology, 15(11): 11451151, https://doi.org/10.1175/1520-0450(1976)015<1145:OMCMTS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gibelin, A.L. and Déqué, M. (2003). Anthropogenic climate change over the Mediterranean region simulated by a global variable resolution model. Climate Dynamics, 20(4): 327339, https://doi.org/10.1007/s00382-002-0277-1.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F. and Lionello, P. (2008). Climate change projections for the Mediterranean region. Global and Planetary Change, 63(2‒3): 90104, https://doi.org/10.1016/j.gloplacha.2007.09.005.

    • Search Google Scholar
    • Export Citation
  • Haslinger, K. and Blöschl, G. (2017). Space-time patterns of meteorological drought events in the European greater alpine region over the past 210 years. Water Resources Research, 53(11): 98079823, https://doi.org/10.1002/2017wr020797.

    • Search Google Scholar
    • Export Citation
  • Held, L. and Sabanés Bové, D. (2020). Markov models for time series analysis. In: Likelihood and Bayesian inference. Statistics for biology and health. Springer, Berlin, Heidelberg, https://doi.org/10.1007/978-3-662-60792-3_10.

    • Search Google Scholar
    • Export Citation
  • Hungarian Central Statistical Office (2016). The World Heritage Tokaj wine region, Available at: https://www.ksh.hu/docs/hun/xftp/idoszaki/pdf/tokaji_borvid.pdf (Accessed 13 October 2023).

    • Search Google Scholar
    • Export Citation
  • HungaroMet (2023). Meteorológiai Adattár (Meteorological Database). Available at: https://odp.met.hu/

  • HungaroMet (Hungarian Meteorological Service Nonprofit Private Limited Company) data (2021a). Available at: https://www.met.hu/eghajlat/eghajlatvaltozas/megfigyelt_hazai_valtozasok/homerseklet_es_csapadektrendek/csapadekosszegek/.

  • HungaroMet (Hungarian Meteorological Service Nonprofit Private Limited Company) data (2021b). Available at: https://www.met.hu/eghajlat/magyarorszag_eghajlata/altalanos_eghajlati_jellemzes/csapadek/.

  • Huo, Z., Mao, H., Yang, J., and Wang, P. (2022). Process-based evaluation indicators of grape drought and risk characteristics in the Bohai Rim Region, China. Theoretical and Applied Climatology, 150(1): 126, https://doi.org/10.21203/rs.3.rs-1724263/v1.

    • Search Google Scholar
    • Export Citation
  • Ibeje, A.O., Osuagwu, J., and Onosakponome, R. (2018). A Markov model for prediction of annual rainfall. Journal of Engineering and Applied Science, 3(11): 15.

    • Search Google Scholar
    • Export Citation
  • Keller, M. (2010). Managing grapevines to optimize fruit development in a challenging environment: a climate change primer for viticulturists. Australian Journal of Grape and Wine Research, 16(s1): 5669, https://doi.org/10.1111/j.1755-0238.2009.00077.x.

    • Search Google Scholar
    • Export Citation
  • Liopa-Tsakalidi, A., Thomopoulos, V., Pantelis, B., and Kavga, A. (2021). A NB-IoT based platform for smart irrigation in vineyard. Conference: 10th International Conference on Modern Circuits and Systems Technologies (MOCAST). https://www.doi.org/10.1109/MOCAST52088.2021.9493381.

    • Search Google Scholar
    • Export Citation
  • Marozzi, M. (2009). Some notes ont the location-scale Cucconi test. Journal of Nonparametric Statistics, 21(5): 629647, https://doi.org/10.1080/10485250902952435.

    • Search Google Scholar
    • Export Citation
  • Marozzi, M. (2013). Nonparametric simultaneous tests for location and scale testing: a comparison of several methods. Communications in Statistics – Simulation and Computation, 42(6): 12981317, https://doi.org/10.1080/03610918.2012.665546.

    • Search Google Scholar
    • Export Citation
  • Martell, D.L. (1999). A Markov chain model of day-to-day changes in the Canadian forest fire weather index. International Journal of Wildland Fire, 9(4): 265273, https://doi.org/10.1071/WF00020.

    • Search Google Scholar
    • Export Citation
  • Medina, A., Gilbert, M.K., Mack, B.M., O’Brian, G.R., Rodríguez, A., Bhatnagar, D., Payne, G., and Magan, N. (2017). Interactions between water activity and temperature on the Aspergillus flavus transcriptome and aflatoxin B1 production. International Journal of Food Microbiology, 256: 3644.

    • Search Google Scholar
    • Export Citation
  • Mesterházy, I., Mészáros, R., and Pongrácz, R. (2014). The effects of climate change on grape production in Hungary. Időjárás − Quarterly Journal of the Hungarian Meteorological Service, 118(3): 193206.

    • Search Google Scholar
    • Export Citation
  • Meyn, S.P. and Tweedie, R.L. (2012). Markov chains and stochastic stability. Springer Science & Business Media, London.

  • Mississippi, V. (2022). Markov chains and applications. Selecciones Matemáticas, 9(1): 5378, https://doi.org/10.17268/sel.mat.2022.01.05.

    • Search Google Scholar
    • Export Citation
  • Mozell, M.R. and Thach, L. (2014). The impact of climate change on the global wine industry: challenges and solutions. Wine Economics and Policy, 3(2): 8189, https://doi.org/10.1016/j.wep.2014.08.001.

    • Search Google Scholar
    • Export Citation
  • Mueller, B. and Seneviratne, S.I. (2012). Hot days induced by precipitation deficits at the global scale. Proceedings of the National Academy of Sciences, 109(31): 1239812403, https://doi.org/10.1073/pnas.1204330109.

    • Search Google Scholar
    • Export Citation
  • Paulo, A.A. and Perreira, L.S. (2007). Prediction of SPI drought class transitions using Markov chains. Water Resources Management, 21(10): 18131827, https://doi.org/10.1007/s11269-006-9129-9.

    • Search Google Scholar
    • Export Citation
  • Pirata, M.S. (2018). Estudo Do Stress Hídrico Da Vinha-Castas Aragonês e Trincadeira [Study of Water Stress in Aragonês and Trincadeira Vineyards], Phd thesis. Universidade de Évora, (in Portuguese).

    • Search Google Scholar
    • Export Citation
  • Ross, S.M. (2010). Introduction to probability models. Academic Press, New York.

  • Szepesi, J., Harangi, Sz., Ésik, Zs., Novák, T.J., Lukács, R., and Soós, I. (2017). Volcanic geoheritage and geotourism perspectives in Hungary: a case of a UNESCO World heritage site, Tokaj wine region historic cultural landscape, Hungary. Geoheritage, 9(3): 329349, https://doi.org/10.1007/s12371-016-0205-0.

    • Search Google Scholar
    • Export Citation
  • Tettey, M., Oduro, F.T., Adedida, D., and Abaye, D.A. (2017). Markov chain analysis of the rainfall patterns of five geographical locations on the south-eastern coast of Ghana. Earth perspectives − Transdisciplinary Enabled, 4(6): 111, https://doi.org/10.1186/s40322-017-0042-6.

    • Search Google Scholar
    • Export Citation
  • Tokhi Arab, S., Noguchi, R., and Ahamed, T. (2022). Yield loss assessment of grapes using composite drought index derived from landsat OLI and TIRS datasets. Remote Sensing Applications Society and Environment, 26(3): 100727, https://www.doi.org/10.1016/j.rsase.2022.100727.

    • Search Google Scholar
    • Export Citation
  • Tomaz, A., Martinez, J.M.C., and Pacheco, C.A. (2015). Yield and quality responses of “Aragonez” grapevines under deficit irrigation and different soil management practices in a mediterranean climate. Ciencia e Tecnica Vitivinicola, 30(1): 920, https://doi.org/10.1051/ctv/20153001009.

    • Search Google Scholar
    • Export Citation
  • Wu, F., Bhatnagar, D., Bui-Klimke, T., Carbone, I., Hellmich, R., Munkvold, G., Paul, P., Payne, G., and Takle, E. (2011). Climate change impacts on mycotoxin risks in US maize. World Mycotoxin Journal, 4(1): 7993, https://doi.org/10.3920/WMJ2010.1246.

    • Search Google Scholar
    • Export Citation
  • Zhang, R., Tang, C., Ma, S., Yuan, H., Gao, L., and Fan, W. (2011). Using Markov chains to analyze changes in wetland trends in arid Yinchuan Plain, China. Mathematical and Computer Modelling, 54(3–4): 924930.

    • Search Google Scholar
    • Export Citation
  • Collapse
  • Expand

Senior editors

Editor(s)-in-Chief: Felföldi, József

Chair of the Editorial Board Szendrő, Péter

Editorial Board

  • Beke, János (Szent István University, Faculty of Mechanical Engineerin, Gödöllő – Hungary)
  • Fenyvesi, László (Szent István University, Faculty of Mechanical Engineering, Gödöllő – Hungary)
  • Szendrő, Péter (Szent István University, Faculty of Mechanical Engineering, Gödöllő – Hungary)
  • Felföldi, József (Szent István University, Faculty of Food Science, Budapest – Hungary)

 

Advisory Board

  • De Baerdemaeker, Josse (KU Leuven, Faculty of Bioscience Engineering, Leuven - Belgium)
  • Funk, David B. (United States Department of Agriculture | USDA • Grain Inspection, Packers and Stockyards Administration (GIPSA), Kansas City – USA
  • Geyer, Martin (Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB), Department of Horticultural Engineering, Potsdam - Germany)
  • Janik, József (Szent István University, Faculty of Mechanical Engineering, Gödöllő – Hungary)
  • Kutzbach, Heinz D. (Institut für Agrartechnik, Fg. Grundlagen der Agrartechnik, Universität Hohenheim – Germany)
  • Mizrach, Amos (Institute of Agricultural Engineering. ARO, the Volcani Center, Bet Dagan – Israel)
  • Neményi, Miklós (Széchenyi University, Department of Biosystems and Food Engineering, Győr – Hungary)
  • Schulze-Lammers, Peter (University of Bonn, Institute of Agricultural Engineering (ILT), Bonn – Germany)
  • Sitkei, György (University of Sopron, Institute of Wood Engineering, Sopron – Hungary)
  • Sun, Da-Wen (University College Dublin, School of Biosystems and Food Engineering, Agriculture and Food Science, Dublin – Ireland)
  • Tóth, László (Szent István University, Faculty of Mechanical Engineering, Gödöllő – Hungary)

Prof. Felföldi, József
Institute: MATE - Hungarian University of Agriculture and Life Sciences, Institute of Food Science and Technology, Department of Measurements and Process Control
Address: 1118 Budapest Somlói út 14-16
E-mail: felfoldi.jozsef@uni-mate.hu

Indexing and Abstracting Services:

  • CABI
  • ERIH PLUS
  • SCOPUS

2023  
Scopus  
CiteScore 1.8
CiteScore rank Q2 (General Agricultural and Biological Sciences)
SNIP 0.497
Scimago  
SJR index 0.258
SJR Q rank Q3

Progress in Agricultural Engineering Sciences
Publication Model Hybrid
Submission Fee none
Printed Color Illustrations 40 EUR (or 10 000 HUF) + VAT / piece
Article Processing Charge 900 EUR/article (only for OA publications)
Regional discounts on country of the funding agency World Bank Lower-middle-income economies: 50%
World Bank Low-income economies: 100%
Further Discounts Editorial Board / Advisory Board members: 50%
Corresponding authors, affiliated to an EISZ member institution subscribing to the journal package of Akadémiai Kiadó: 100%
Subscription fee 2025 Online subsscription: 173 EUR / 190 USD
Print + online subscription: 200 EUR / 220 USD
Subscription Information Online subscribers are entitled access to all back issues published by Akadémiai Kiadó for each title for the duration of the subscription, as well as Online First content for the subscribed content.
Purchase per Title Individual articles can be purchased at the prices indicated.

Progress in Agricultural Engineering Sciences
Language English
Size B5
Year of
Foundation
2004
Volumes
per Year
1
Issues
per Year
1
Founder Magyar Tudományos Akadémia  
Founder's
Address
H-1051 Budapest, Hungary, Széchenyi István tér 9.
Publisher Akadémiai Kiadó
Publisher's
Address
H-1117 Budapest, Hungary 1516 Budapest, PO Box 245.
Responsible
Publisher
Chief Executive Officer, Akadémiai Kiadó
ISSN 1786-335X (Print)
ISSN 1787-0321 (Online)

Monthly Content Usage

Abstract Views Full Text Views PDF Downloads
Sep 2024 0 1489 75
Oct 2024 0 1583 27
Nov 2024 0 665 26
Dec 2024 0 97 22
Jan 2025 0 289 30
Feb 2025 0 154 7
Mar 2025 0 0 0