The change in weather types in two population centers in northwestern Iran (Urmia and Tabriz) around Lake Urmia with the Woś classification approach Open Access

Climate change refers to statistically significant changes in the average state of the climate or its variability over a long period of time (usually decades or more). Climate change may be due to natural internal processes or external forcings, or due to ongoing anthropogenic changes in atmospheric composition or land use. The change in the extreme and average values of climatic parameters is one of the important consequences of climate change (IPCC 2014). Moreover in recent years, due to the harmful economic, social and financial consequences related to atmospheric phenomena, the issue of climate change has become very important. Some factors of climate change may be unknown, and it cannot be considered as the result of human activities alone; however, the human interventions with nature have accelerated the climatic changes and its consequences, especially in recent decades, as evident from several studies (Guptha et al. 2021, 2022; Swain et al. 2021, 2022a, 2022b; Sahoo et al. 2022; Nandi & Swain 2023). These effects have gone so far that researchers have substantiated the impact of climate change on migration (Moniruzzaman et al. 2018), food security (Ghalehsard et al. 2021), agricultural and horticultural production (Heydari & Movaghari 2021), geopolitics (Yang et al. 2020), resources water and health (Kumar 2021), security (Brown et al. 2007), drought (Gleick 2014,) human rights (Levy & Patz 2015) and extreme precipitation (Javan et al. 2023). In the meantime, although there are many challenges in the application of different models in climate change, but so far, various methods such as general circulation models of the atmosphere such as GCMS and regional climate models (RCMS) have been used (Prudhomme & Davies 2008; Hagemann et al. 2009), synoptic analyzes (Fazel-Rastgar 2021), experimental methods such as sediment analysis (Van der Linden et al. 2008) and investigation of plant growth rings (Williams et al. 2010), GIS methods (Kundu & Dutta 2011), and Statistical methods such as Kendall and Sen's slope estimator tests (Ashraf et al. 2021) have been used. Based on this, the climate data used based on the targeting of the researcher mainly relies on one or two or more variables, which are mostly single. Considering that all climatic variables act in a system format in the atmosphere, therefore, in the study of climatic changes, to obtain better results, it is necessary to use multiple and simultaneous climatic variables as much as possible. On the other hand, it should also be noted that the climate structure of each place is built based on the frequency of the weather types that occur there and over time. Based on this, each type of weather is a weather with certain characteristics, which if it occurs repeatedly in a place, will be the dominant weather of that place. (Allijani & Kavyani 2019). Therefore, the climate of a place is the result of the effect of weather types in a place for a long time, which has a significant impact on human activities and its plant and animal community, and any change in its characteristics can affect them. In this regard, Complex climatology methods are one of the proposed methods for determining the characteristics of weather, and various researchers have published various articles on the impact of weather on life, economy and energy (Ferdynus 2012; Hidalgo & Jougla 2018). Although the term Complex Climatology was first proposed by Fedorov in 1921 in Russia and then developed in Germany and the United States with a completely objective approach based on the use of data obtained from observations in one place under the title of weather types (Court 1957), But it must be said that the invention of methods that can combine the characteristics of the air in the form of a type of weather has a long history. (Błażejczyk 1985) In this regard, various classification methods have been devised based on the experimental framework based on selected values of meteorological elements according to the determined ranges, and genetics based on synoptic characteristics or types of weather circulation in a place. So far, various researchers in the world have used the experimental method to determine whether types based on the use of selected meteorological elements, which in recent decades can be referred to as (Hufty 1971) for Canada, (Besancenot et al. 1978) for areas of the Iberian Peninsula, (Alcoforado et al. 2004) for North America, (Andrade et al. 2007) for western Portugal, (Bielec-Bakowska & Piotrowicz 2011) for Krakow, Poland (Cantat & Savouret 2014), for Southeast China (Wang & Sun 2021), (Piotrowicz & Ciaranek 2021) for Poland. The review of the history of the conducted research shows that after years of presenting articles on the determination of weather types based on the general circulation of the atmosphere, studies based on Complex climatology have received attention (Philipp et al. 2010). One of the proposed methods in recent years is Woś classification, which was invented between 1968 and 2010 and is used in this article (Cantat & Savouret 2014), but they have been used in bioclimatic and bio meteorological studies, nature tourism, agricultural climatology and about plant pests.

In recent years, Lake Urmia, as the largest permanent lake in the Middle East and the second saline lake in the world, has faced one of the tragic environmental crises by losing a large part of its area, and in the past two decades, the water level has decreased by about 8 meters (Khazaei et al. 2019; Valiallahi et al. 2019). Although the role of destructive human interference in the occurrence of this crisis cannot be hidden, one of its factors is climate change (Bashirian et al. 2020). Certainly, due to the dryness of the lake Urmia and the fact that lakes play a role as guardians of climate change (Huang et al. 2023), the climatic patterns of the region have undergone drastic changes with the change of weather types, and the permanent population centers, especially the two densely populated centers studied in this article, have had an impact. In this field, most of the studies carried out by researchers focus on the use of changes in one or more climate elements separately (Modarres & Sarhadi 2009; Vaheddoost & Aksoy 2017; Dehghanipour et al. 2020), changes related to drought indices (Abbasian et al. 2021), Changes related to water resources (Alizadeh Govarchin Ghale et al. 2018) or economic (Schmidt et al. 2021), changes leading to the occurrence of dust storms (Gholampour et al. 2015), climate changes according to climate scenarios (Zohrevandi et al. 2020) and Jani et al. 2023) and synoptic (Hanafi 2017). Based on this, the goal of this article is to use a multi-element method of meteorological variables based on coding (Woś classification) for the possibility of numerical analysis and processing, so that the process of their changes and the dominance of new types according to Complex climatology can be discovered in the study area. Most of the works done by researchers in Iran for weather classification in different regions of the country are based on the synoptic climatology method (Alijani 2010; Tavusi et al. 2010 and Khosravi & Nazaripour 2013). Therefore, this research tries to determine the weather types in Urmia and Tabriz stations located in the west and east of Lake Urmia with the Woś classification approach and to discover the resulting changes during the time of the drying tragedy of this lake based on the modified Mann Kendall (MMK) method.

A general survey of the climate of the two stations, despite their not-so-great distance, indicates obvious differences. Based on this, as seen in Table 1, the average minimum temperature range in Urmia is lower than Tabriz, but the difference in minimum temperature range in both stations is about 4 °C. Meanwhile, the absolute minimum temperature values in these two stations are slightly different. The same situation can be seen in the average maximum temperature, but the difference in the range of absolute maximum temperature values reaches 6°. The average temperature also shows slight changes, but it is higher in Tabriz. Regarding the precipitation, there is a big difference between the two stations, and the comparison of the minimum and maximum precipitation values recorded during the statistical period indicates extreme changes and fluctuations. Examining the percentage of cloudiness also shows that despite its high values in Tabriz, it has practically no effect on the precipitation due to the lack of necessary conditions for precipitation. Of course, it should be pointed out that the variety of geographical factors of the region such as Lake Urmia and the diverse combination of climatic parameters have made the climatic conditions such that despite the proximity of the distance and small differences between the parameters, different weather types occur in these two stations.

As mentioned before, this article is based on the Woś classification method (1999 and 2010) and using the daily data of air temperature (maximum, minimum and average temperature in degrees Celsius), precipitation (in mm) and percentage of cloud cover (Table 2). This method is one of the interesting methods based on data processing, which has been especially noticed in Poland (Bielec-Bakowska & Piotrowicz 2011). Of course, in some cases, by changing this method, for example, by adding an additional element such as wind speed (Ferdynus & Marsz 2000) and by changing the range of elements mentioned above, they seek a new method for classifying types of weather in areas such as poles. In this classification, Woś has classified the values of the above-mentioned climatic elements in 16 domains (12 for temperature, 3 for cloudiness, 2 for precipitation) and has classified the weather in the form of 66 types. The types of weather are coded with numbers. In this way, the first digit up to the number 9 (or the first two digits of the frost weather type with digits 10 to 12) shows the thermal characteristics, the next (penultimate) digit shows the cloudy characteristic and the last digit also shows the precipitation characteristic. In total, 12 weather types are separated based on temperature, which can be classified into three groups (Table 2): T_mean, T_max, T_min > 0° (with symbol 1–5), Frosty T_max > 0°, T_min < 0°) with symbol (6–9), Freezing (T_max and T_min < 0° with the symbol (10–12). The other group is classified with weather subtype, which carries information about cloud cover and daily precipitation. In this way, each weather type is obtained with a three- or four-digit code with an underscore. In the classification of Woś (1999), the air elements used in the classification to create weather groups are linked with the coding method, while some classification methods lack such a capability. This link between the elements with a hyphen ‘-’ is such that, for example, the type of weather specified as 20-3 indicates warm weather with an average daily temperature of 10–15 °C and a maximum temperature of more than 0 °C with cloudy skies. It is cloudy with more than 80% and no precipitation.

Table 2

Classification of weather types (Woś 1999)

View Large

The Mann–Kendall (MK) test is one of the non-parametric statistical test methods proposed by the World Meteorological Organization as a standard method for climate change analysis (Mitchell et al. 1966; Kundzewicz & Robson 2000). This test is used for data with a trend and is based on a rank process, and it can be used in data series that do not have a specific statistical distribution and contain residual values or non-linear trends (Karpouzos et al. 2010). However, in real conditions, some hydrological and climatological time series have autocorrelation coefficients. Autocorrelation affects the amount of variance calculated in the MK test, and in this case, although the data lacks a trend, a trend is detected for them with the MK method (Bagherpoor1 et al. 2017). In other words, in such a data time series, neither the shape of the distribution nor the mean of the MK test statistic has changed, but its variance changes (its positive autocorrelation increases and vice versa). This led to the development of new methods based on the correction of the variance of the MK statistic. Researchers have proposed various methods to eliminate autocorrelation, under the name of modified Mann–Kendall (MMK) test (Hamed & Ramachandra Rao 1998; Yue & Wang 2002, 2004). Therefore, researchers prefer MMK over MK (Kocsis & Anda 2018; Nazeri Tahroudi et al. 2019; and Swain et al. 2022a, 2022b). In this research, the MMK test was used using the variance correction approach of Hamed & Ramachandra Rao (1998). On this basis, after correcting the variance and removing the effect of autocorrelation, the amount of occurrence of any significant trend is determined by the normal approximation and the value of Zc (corrected Z value). If Zc values are positive, it indicates an upward trend, and negative values indicate a downward trend, and Tau indicates the Menn-Kendall correlation coefficient between the data and time. The Sen's slope estimator test is a type of linear trend analysis test and determines the slope of changes with a confidence interval.

This test is represented by Q value (Sen 1968). In recent years, the above tests have been widely used by various researchers in Iran and have shown good results in accordance with the climatic reality of the region under study. In this research, weather types in two selected stations have been analyzed in the R program by the above two statistical tests. At the same time, Sequential Mann–Kendall (SQMK) test was used in this research. This method is used to determine the start time of the trend (jump points) within the data series (Sneyres 1990). In fact, this method calculates the statistical values in the time series of the data with the MK ranking method. The opposite is calculated; That is, it is assumed that the end of the series is at the beginning. When the trend is acceptable at a significant level, the change point can be determined with the help of the Man-Kendall chart. In the next step, to determine the jump points of the sequence U and U’ based on time (i), draw a graph. In the significant state of the trend, the two graphs at the starting point of the phenomenon will cross each other outside the range of ±1.96 and move in the opposite direction. This point is called the jump point collision (Sabziparvar et al. 2011). In the case that the series is stationary, the two sequences U and ‘U will act in parallel or collide several times so as to change the jump. Based on this, the placement of the curve above the range of +1.96 indicates an upward trend and its location below the range of −1.96 indicates a downward trend.

Also, the figure shows that since 2000 (before the water level of Lake Urmia decreased drastically) weather types in Tabriz have lost their stability and suffered some kind of chaos.

On the other hand, the examination of Table 4 also shows that despite the presence of 3 climatic elements used, air temperature is still the most important factor for differentiating weather types. As from the 8 types of weather types, 61.7% in Urmia and 61.4% in Tabriz have been allocated to them.

Also, Table 5 was prepared to determine the most occurrence of monthly weather types. In both stations, regardless of the coordination in the occurrence of the main types, the main difference in cloudiness and precipitation is completely observed, and the percentage values between Urmia and Tabriz show a fundamental difference.

Regarding the evaluation of change of weather types according to Table 6, The trend of changes in the main weather types in Tabriz indicates a significant positive trend of types 1 (hot weather type) and 6 (slightly cold weather type) and a significant negative trend of type 10 (slightly frosty weather type) at the 95% confidence level (P-values <0.05 are highlighted in bold and underlined). The results of both the above trends indicate an increase in air temperature toward warm types. Meanwhile, in Urmia, only type 1 and type 6 show a significant positive trend. The examination of the Q statistic (Sen's slope estimator) shows the greatest increase in both stations in type 1.

Examining the trend of changes in subtypes (Table 7) in Tabriz shows a significant positive trend in subtype _10 (cloudy without precipitation), and a negative trend in subtype _20 (very cloudy without precipitation). In Urmia, in addition to the above two subtypes, subtype 21 (very cloudy with rain) shows significant changes at the 95% confidence level (P-values <0.05 are highlighted in bold and underlined). These conditions show the trend of decreasing precipitation, even in cloudy conditions above 80%. In addition, the examination of the negative trend of the subtypes also indicates a little cloudiness and precipitation. In other words, cloudiness and precipitation in both regions with a negative trend (of course, according to the age statistics, it is more intense in Tabriz) has caused a decrease in receiving adequate precipitation.

On the other hand, in order to identify the most changes in the subtypes of weather during different months of the year, Table 8 was prepared. It should be noted that due to the size of the prepared table, only significant values (P-value < 0.05) are mentioned here. Accordingly, in Tabriz, subtype _10 (cloudy without precipitation) shows a positive trend in March and August, _00 (sunny without precipitation) shows a positive trend in February, and _11 (cloudy with precipitation) shows a positive trend in September. Meanwhile, the subtypes _20 (very cloudy without precipitation) in February, _21 (very cloudy with precipitation) in March and _00 (sunny without precipitation) in August show a negative trend. The general pattern of this trend is the tendency to increase without precipitation even with the presence of clouds in Tabriz. This situation has continued to occur in Urmia in more types during the months of the year, especially in the winter and autumn seasons. The above investigations show that firstly, the approach of the changes in both locations toward the subtypes without precipitation is changing even with the presence of cloudiness and secondly, the comparative examination of the Sen's slope estimator values in the subtypes and the same months indicates a higher relative intensity of these changes. At the same time, the aforementioned changes are important from the point of view that most of these changes are taking place in the winter and autumn seasons, which are important rainy seasons in Tabriz and Urmia.

The purpose of this study was to investigate the weather types in two stations on the sides of Lake Urmia, considering the west-east air currents, based on the Woś classification method and how it trended over the years. In this regard, according to the capabilities of the Woś method, first It provides the possibility of identifying weather types as a combination of climatic elements, and secondly, it makes possible the daily data analysis of these elements during the long-term statistical period in small amounts, it has a special place, because based on this, numerical and statistical analyzes are possible can be done on the resulting data. Due to the possibility of using long-term statistics in this method, it is possible to provide a better climate view than the types of weather types obtained through synoptic and even compared to classical climatology that climate It expresses each place with the average value of each meteorological element, it provides a better possibility to fully determine the weather condition (weather type) (Piotrowicz 2010).

Based on this, most of the presented classification methods such as Köppen, De Martonne and Thornswaite are placed in the framework of classical climatology. If the method presented in this article, which is in the framework of complex climatology, by calculating the frequencies of weather types, shows a more accurate climate reality than the average of meteorological parameters or single numerical indices that are calculated from meteorological averages (Alcoforado et al. 2004).

Therefore, the investigations revealed that although the first weather type starts later in Tabriz, the frequency of its occurrence is actually much higher than in Urmia. Meanwhile, the second weather type has a higher frequency. In this regard, the trend of increasing type 1 in both stations and of course in Tabriz is changing more intensively, and in other words, the general trend indicates an increase in the intensity of heating of the hot weather type at the time of its occurrence, i.e. in the summer season. At the same time, the examination of the subtypes also indicated the predominance of the subtype of sunny weather without rain and cloudy weather without rain in most of the year, especially in summer, but the variety of types is relatively greater in other seasons. Darand et al. 2017 also refer to similar types of this type in their study. At the same time, despite the mountainous location of these two population centers in the northwest of Iran, it can be seen that although there is more variety in the weather types in the cold seasons, the occurrence of the expected cold types has decreased during the statistical period studied. As type 6 has an increasing trend in both places, and with the increase in the occurrence of other colder types, it has decreased. This condition is also more severe in Tabriz compared to Urmia. At the same time, the general pattern indicates a decrease in the variety of weather types toward the end years of the period is studied. Considering the weather subtypes, that means the amount of cloudiness and precipitation, it was determined that subtype _00 (sunny without precipitation) had the most occurrences. Of course, Urmia has the highest percentage. In the next place, subtype _10 (cloudy without precipitation) among which Tabriz has the highest percentage. The change trend of this subtype is also increasing and it is very intense in Urmia. The importance of the above problem becomes complicated when we know that the total rainfall Urmia and Tabriz occur continuously between October and May, and according to Table 3, subtype _10 (cloudy without precipitation) has an increasing trend during this period, and even subtypes _20 (very cloudy without precipitation) and _21 (very cloudy with precipitation) as well have a negative trend, which indicates the general transformation of the types related to precipitation in both places. Of course, considering Table 6, the general changes of the types indicate the decreasing trend of precipitation in the above months. The results are found to be consistent with the previous studies (Zarghami et al. 2009; Delju et al. 2013; Jalili et al. 2016; Alizadeh Govarchin Ghale et al. 2018; Arkian et al. 2018; Chaudhari et al. 2018; Pooralihossein & Delavar 2020). The results of this research are found to be consistent with the study of Javan et al. (2023), who investigated the future changes in precipitation extreme indices in the Lake Urmia Basin during the period 2021–2100 compared to the base period (1987–2016), using the CMIP5 models, which showed that the average precipitation of the basin will decrease by the end of the 21st century.

Another point is the approximate correspondence of the changes in weather patterns and especially the increase in the hot air pattern (1) with the changes in Lake Urmia, which indicates the beginning of the increase in the intensity of this pattern with its drying process, which according to reports started in the decade of 2000 and its effect on Tabriz is completely visible. At the same time, the decrease in the variety of weather types during the statistical period under review indicates the stability of the weather toward warm and hot types. Also, some researchers believe that one of the main factors in the drying up of Lake Urmia is the overuse of water resources in the region, and this excessive use causes the aggravation of climate changes in Lake Urmia and the cities of Urmia and Tabriz, and changes the types of the weather. Although this research is based on the daily data of ground stations, but its results are in high agreement with the studies that have been done from the synoptic point of view in the region (Karimi et al. 2018). Based on the findings and their comparison with the studies conducted in Iran and globally, it seems that the studied population centers are exposed to climate change. The change in weather patterns during the study period indicates this. Of course, as mentioned before, since the drying up of Lake Urmia, the intensity of these changes has increased, and in the meantime, in the east of the lake (Tabriz), the effects immediately showed. While, in the west of it (Urmia), with a period delay, the trend of changes was observed. SQMK shows this perfectly. This issue is important because Rousta et al. (2023), by examining the changes in the levels of local lakes in the countries of Iran, Iraq, and Turkey and comparing the changes, they have come to the conclusion that the severity of the drying of Lake Urmia, regardless of natural factors, is more severe due to destructive human interference and somehow leads to climate change. On the other hand, the changes occurred especially in the autumn and winter seasons of the region and somehow warmed these seasons compared to normal conditions. This issue is also important from the point of view of knowing that the two mentioned seasons are the most important rainy seasons and water resource storage in the region, which due to the new conditions of cloudiness and low rainfall, can be a threat to the future of agriculture and water resources in the region and somehow the development and progress of not only the two studied population centers but also other neighboring areas. According to the nature of this research, which is based on data processing and numerical analysis of weather types in daily periods to long periods of time, it is possible to help policymakers in the process of restoring the lake and curbing its drying, of course, by studying a larger number of stations, especially in urban areas, along with other methods.

This study, using the data of Urmia and Tabriz meteorological stations located on both sides of Lake Urmia in the northwest of Iran, based on the Woś classification method, which is based on the Complex climatology perspective and can be used on small scales (unlike synoptic climatology) weather types were extracted and analyzed. The examination of the main weather types in these two important population centers in the northwest of the country indicates the predominance of hot and very hot weather types in summer, especially in Tabriz, and cool to cold and very cold weather types in Urmia. Weather subtypes mainly in both stations include clear, cloudy or cloudless sky subtypes with no precipitation and cloudy sky subtypes occur with precipitation in the cold season. A general survey of weather types shows the greatest diversity in cold seasons in both stations. Examining the trend of changes in weather types also indicates the positive trend of hot and slightly cold weather types, as well as the negative trend of colder types in favor of the increase of cool types in autumn and winter and hot in spring and summer. These conditions have even increased the subtypes of cloudy skies without precipitation, especially in the rainy season in Tabriz. This situation cannot be unrelated to the drying up of Lake Urmia and the increase in the amount of solar energy received, and as a result of intensifying global warming and creating an obstacle in the precipitation process. This condition in Tabriz coincides with the drying up of Lake Urmia since 2000. It is obvious that the continuation of hot weather types can cause many problems in the field of water resources and agriculture, tourism calendar, energy consumption, migration, emptying of villages and settlements in cities and some other problems by increasing the intensity of evaporation in the region. Today, the environmental effects of climate change have gone beyond its obvious consequences and have entered important aspects of human life, including politics. It seems that the policymakers by adopting policies such as architecture compatible with the climate in cities in order to use less fossil fuels, cultivate species adapted to the new climatic conditions, optimal management of water and irrigation resources, monitoring the non-expansion of agricultural lands and other methods in accordance with local conditions, they can control the speed and intensity of changing weather patterns and finally prevent its destructive effects to a large extent.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.