Abstract

The time-placement scheme of climate extreme changes is important. In this regard, a set of a compound indices derived using daily resolution climatic time series data is examined to assess climate change in Iran. The compound indices were examined for 47 synoptic meteorological stations during 1981–2015. The results show that most stations experienced a negative trend for the cool/dry (CD) and cool/wet (CW) index and a positive trend in CW was observed in some dispersed small areas. Both warm/dry (WD) and warm/wet (WW) indices have similar behavior, but the magnitude and spatial consistency of WW days were much less than WD days. The results show that more than 80% of stations experienced a decrease in the annual occurrence of the cold modes and an increase in the annual occurrence of the warm modes. On the other hand, universal thermal climate index (UTCI) change demonstrated a significant increase in the annual occurrence of strong heat stress (32–38 °C) and significant decrease in the annual occurrence of no thermal stress class (9–26 °C). Moreover, trends in tourism climate index (TCI), including TCI 60 and TCI 80, showed similar changes but with weak spatial coherence.

INTRODUCTION

Climate change is considered as a global challenge whose effects differ by region and local characteristics (Leonard et al. 2014; Filho et al. 2016). During the past decades, large positive temperature increases from their mean values have become common in many parts of the world. In both hemispheres, the land surface temperature has increased at a faster rate than the oceans according to the Intergovernmental Panel on Climate Change (IPCC 2007). The Earth's surface temperature has maximized over the last three decades and has seen increasing trends since 1850. In this regard, the northern hemisphere has become much warmer from 1983 to 2012 than any 30-year period in the last 1,400 years. The linear trends of the average land and ocean surface combination temperature data during the period 1880 to 2012 reveal an increase of about 0.85 °C (IPCC 2013). The planning and performance of human community infrastructures in relation to food, water, energy, shelter, and transportation depend on climate extreme change values according to the World Meteorological Organization (WMO 2009). The economic changes and the effects of extreme climate change are among the most important challenges humans face regarding climate change, states the Climate Change Science Program (CCSP 2008). The latest outputs of meteorological models suggest that global warming has caused an increase in evaporation from land surface and surface water bodies, which is anticipated to have a serious impact over time on water resources management and the global population (Moazenzadeh et al. 2018). Meanwhile, global warming has caused increasing and several climate events such as drought, heat waves, floods, and fires in different parts of the world (Alexander et al. 2006; Aghakouchak et al. 2014; Leonard et al. 2014). Predicting climate extremes based on temperature and precipitation index in the Fifth Report of the IPCC highlights a significant growth in the number and magnitude of warm and dry periods (Hao et al. 2013; Filho et al. 2016). The daily index consists of temperature or precipitation extremes which have been used for assessing long-time variability and frequency of days above or below specific physically based thresholds (Zhang et al. 2011). Mainly, due to the interdependence and thermodynamics, relations between precipitation and temperature have been addressed in numerous studies (Liu et al. 2012). The main goal of analyzing extremes is to describe the balance and changes of climate to find the most desirable balance in macro plans for applying high safety standards and preventing great loss to communities and systems with regard to severe events. Sometimes, one variable may be in an extreme state, but more often, these events can be considered as a combination of variables, not all of which are necessarily extreme (Leonard et al. 2014). An extreme impact may be a combination of one or more variables leading to a severe change in climate, referred to as a compound event. According to copula theory (Miao et al. 2016), recent studies have been conducted based on describing the compound index of climate combinations as warm/dry, warm/wet, cool/dry, and cool/wet (Lopez-Moreno et al. 2011; Estrella & Menzel 2012; Arsenovic et al. 2013; Hao et al. 2013). These compound indices reveal the change at all spatial or temporal scales with significant trends in the frequency of the cool and warm modes.

Most parts of Iran are located in the desert belt, which is affected by extreme climate change and trends. Therefore, considering trends and climate change and their effects is essential and unavoidable for planning risk management in relation to communities, systems, and infrastructures through adaptation and the bad effects of climate change. Several national studies have investigated trends of climatic variables in Iran. In recent studies, several extreme precipitation and temperature indices have been analyzed (Rahimzadeh et al. 2009; Tabari et al. 2011; Parak et al. 2015), but the climate change was not concerned with the use of compound extreme indices. These studies provided important findings, although they have focused on only the individual elements of the climate and have used mean and extreme values of climate variables. This paper is one of the first national studies to be based on compound values. In this study, the composition of more different climatic elements is considered and several statistical tests were used for trend detection. The main purpose of this paper is to demonstrate the trend of compound indices in order to find the impact of climate change over the study area. In this regard, the trends of combined temperature, precipitation, humidity, wind speed, and sunshine statistics in a spatial domain in Iran are investigated as the main research objectives. The aim of the present research was to understand the variations in compound values of temperature and precipitation and other variables. Compared to previous studies, this can provide a comprehensive view of the behavior of the cool and warm modes of heat and moisture, of which analysis of the statistics of each variable is taken individually.

DATA AND METHODS

Iran is the spatial domain used in this study (latitude 25–40 and longitude 44–64). The total area is 1,648,000 km2 and it includes a population of about 75 million, according to the latest National Census of Iran. The climate of Iran varies due to differences in latitude, altitude, and a range of geographic features, including mountains and deserts. Currently, daily data are available for more than 150 synoptic stations in Iran, but there are long-term records for only a few stations. Most of the stations, especially in the earlier years, contain inhomogeneities and uncertainties in their data sets. Our study was limited to only 45 stations, due to the problems mentioned above, or to the presence of wide data gaps, and also to the shortness of record length. On the other hand, the data after 2015 have either been unverified or are not available. The selected stations take into account the length and completeness of records. In order to interpret the climate change compound index, the data were collected from 45 synoptic stations in Iran for the period 1981–2015. These stations have long-time data and are well distributed in different climate regions and elevation levels. The data have been obtained from Iran Meteorological Organization (IRIMO) (Figure 1 and Table 1).

Table 1

Summary of the characteristics and details of the selected synoptic stations

StationHeight (m)Type of climateStationHeight (m)Type of climateStationHeight (m)Type of climate
Abadan 6.6 BWh Bushehr BSh Shahrud 1,325.2 BWk 
Arak 1,702.8 DCsa Tabriz 1,361 BSk Shahrekord 2,048.9 DCsa 
Ahvaz 22.5 BSh Tehran 1,191 BSh Shiraz 1,488 Csa 
Esfahan 1,550.4 BWh Torbat-e Heydariyeh 1,451 Doa Qazvin 1,279.1 Doa 
Orumiyeh 1,328 DCsa Khorramabad 1,147.8 Csa Gorgan Csa 
Bandar-e-Anzali −23.6 Cfa Khoy 1,103.4 BSk Kerman 1,754 BWh 
Babolsar −21 C Ramsar −20 Cfa Kermanshah 1,318.5 Doa 
Bam 1,066.9 BWh Rasht −8.6 Cfa Mashhad 999.2 BSk 
Bandarabbas 9.8 BWh Zahedan 1,370 BWh Yazd 1,230.2 BWh 
Birjand 1,491 BSh Zabol 489.2 BWh Hamedan 1,679.7 DCsa 
Khodabandeh 1,659.4 DCsa Hamedan 1,740.8 DCsa Fasa 1,268 Csa 
Sabzevar 972 BSh Ardebil 1,335.2 DCsb Esfahan 1,551.9 BWk 
Sanandaj 1,373.4 Doa Nowshahr −20.9 Cfa Abadeh 2,030 BSk 
Saghez 1,522.8 DCsa Semnan 1,127 BWh Kish Island 30 BWh 
Chahbahar BWh Kashan 955 BWh Bandar-e-Lengeh 22.7 BWh 
Bojnurd 1,065 BSk Iranshahr 591.1 BWh    
StationHeight (m)Type of climateStationHeight (m)Type of climateStationHeight (m)Type of climate
Abadan 6.6 BWh Bushehr BSh Shahrud 1,325.2 BWk 
Arak 1,702.8 DCsa Tabriz 1,361 BSk Shahrekord 2,048.9 DCsa 
Ahvaz 22.5 BSh Tehran 1,191 BSh Shiraz 1,488 Csa 
Esfahan 1,550.4 BWh Torbat-e Heydariyeh 1,451 Doa Qazvin 1,279.1 Doa 
Orumiyeh 1,328 DCsa Khorramabad 1,147.8 Csa Gorgan Csa 
Bandar-e-Anzali −23.6 Cfa Khoy 1,103.4 BSk Kerman 1,754 BWh 
Babolsar −21 C Ramsar −20 Cfa Kermanshah 1,318.5 Doa 
Bam 1,066.9 BWh Rasht −8.6 Cfa Mashhad 999.2 BSk 
Bandarabbas 9.8 BWh Zahedan 1,370 BWh Yazd 1,230.2 BWh 
Birjand 1,491 BSh Zabol 489.2 BWh Hamedan 1,679.7 DCsa 
Khodabandeh 1,659.4 DCsa Hamedan 1,740.8 DCsa Fasa 1,268 Csa 
Sabzevar 972 BSh Ardebil 1,335.2 DCsb Esfahan 1,551.9 BWk 
Sanandaj 1,373.4 Doa Nowshahr −20.9 Cfa Abadeh 2,030 BSk 
Saghez 1,522.8 DCsa Semnan 1,127 BWh Kish Island 30 BWh 
Chahbahar BWh Kashan 955 BWh Bandar-e-Lengeh 22.7 BWh 
Bojnurd 1,065 BSk Iranshahr 591.1 BWh    
Figure 1

Geographical positions of the 47 synoptic stations over Iran.

Figure 1

Geographical positions of the 47 synoptic stations over Iran.

Methodology

Data quality and homogeneity

Due to the importance of climate change studies, the WMO focuses on quality control and data homogeneity (WMO 2009). One of the most important factors of data inhomogeneity is the change in observation, including time of observation and tools, change of station location, formula, and change in data processing (WMO 2011). Studies of extreme index in climate change have been carried out based on meteorological observation in Iran. As they have errors and uncertainty (Rahimzadeh et al. 2009), so the climatic series have inhomogeneity with sudden and abnormal spikes and deviations and unreal interpretation of climate change. The objective of quality control is to verify whether a reported data value is representative of what was intended to be measured and has not been contaminated by unrelated factors. In this study, ClimPACT software, available as Free Software, is undertaken as the main tool for data analysis. This software is written in R, a language and environment for statistical computing and graphics, and developed by the WMO Expert Team on Climate Change Detection and Indices. First, the ClimPACT quality control (QC) revealed unreasonable values of temperature and precipitation data, such as daily precipitation amounts less than zero and daily maximum temperature less than daily minimum temperature. In addition, the QC also identified outliers in daily maximum and minimum temperatures. It is worth mentioning that all potential data can be evaluated by information on the next and previous day of the event with specialized knowledge about local conditions. The QC test of other climatic variables such as humidity, sunshine, and wind speed was done using simple statistical methods in SPSS software.

There are several methods for the assessment of homogeneity in time series data including relative and absolute methods. The relative tests are standard normal homogeneity test and, more reliably, it is recommended that the test is conducted by correlating the test data series with the homogeneous data series of a neighboring station (Tayanç et al. 1998; Wijngaard et al. 2003; Tsidu 2012). The research uses the R homogeneity test (RH-test) V5 software (Wang & Feng 2013) by using the penalized maximal F-test (PMF) methods to check the homogeneity of the monthly temperature and precipitation and other variables' meteorological data. The software method is based on the PMF or F-test and can identify, and adjust for, multiple change points in a time series (Wang 2008). The F-test is a relative test for structural breaks with reference series, namely multiple change point detection method, and the RH-test has the best detection rate (Yozgatligil & Yazici 2015). In the case of a lack of reference station, the F-test is used for inhomogeneity detection in time series (Wang et al. 2007; WMO 2011; Vincent et al. 2012; Fan & Chen 2016). PMF requires the use of reference stations for the homogeneity analysis, but PMF can be used as an absolute method (i.e., when there are no neighbouring stations to use for comparison). According to the diagrams, if any heterogeneous factors are identified, the time series is assumed to be inhomogeneous, otherwise the time series is homogeneous.

Compound index

The aim of the indicators is to illustrate the temporal and spatial distribution of climate change. It is important to develop a set of compound indices that are statistically robust to cover a wide range of climates and detect changes in climate extremes. Combined indices and heat wave indices are a set of statistical indices that can cover a wide range of climate characteristics and detect variability and changes in climate. In other words, most extreme weather events are the result of combining climate variables. For example, high temperatures coupled with reduced rainfall can cause heat waves and droughts (Klok & Klein Tank 2008). Therefore, following annual study, a specified compound extreme index consisting of seven indices was considered as specified in Table 2.

Table 2

List of the ECA&D (2013) climate index

No.IndexDefinitionUnit
CD Days with TG < 25th percentile of daily mean temperature and RR < 25th percentile of daily precipitation sum (cold/dry days) Days 
CW Days with TG < 25th percentile of daily mean temperature and RR > 75th percentile of daily precipitation sum (cold/wet days) Days 
WD Days with TG > 75th percentile of daily mean temperature and RR < 25th percentile of daily precipitation sum (warm/dry days) Days 
WW Days with TG > 75th percentile of daily mean temperature and RR > 75th percentile of daily precipitation sum (warm/wet days) Days 
UTCI Mean of the universal thermal climate index °C 
TCI60 Days where the tourism climatic index _ 60 Days 
TCI80 Days where the tourism climatic index _ 80 Days 
No.IndexDefinitionUnit
CD Days with TG < 25th percentile of daily mean temperature and RR < 25th percentile of daily precipitation sum (cold/dry days) Days 
CW Days with TG < 25th percentile of daily mean temperature and RR > 75th percentile of daily precipitation sum (cold/wet days) Days 
WD Days with TG > 75th percentile of daily mean temperature and RR < 25th percentile of daily precipitation sum (warm/dry days) Days 
WW Days with TG > 75th percentile of daily mean temperature and RR > 75th percentile of daily precipitation sum (warm/wet days) Days 
UTCI Mean of the universal thermal climate index °C 
TCI60 Days where the tourism climatic index _ 60 Days 
TCI80 Days where the tourism climatic index _ 80 Days 

All indices are calculated annually.

As illustrated in Table 2, the combination of the precipitation and temperature quantities represent the four climate combinations: cool/dry (low temperature and low precipitation), cool/wet (low temperature and high precipitation), warm/dry (high temperature and low precipitation), and warm/wet (high temperature and high precipitation). The 10 and 90 percent quantities define an extreme event, according to an IPCC report (2007) and the 25% and 75% values of precipitation and temperature are used as threshold levels for defining the compound extremes of temperature and precipitation in order to capture a large number of events.

Another compound index, which was used in this study, is the UTCI. The scale of the index is able to express even slight differences in the intensity of meteorological stimulus. The assessment of the thermo-physiological effects of the atmospheric environment is one of the key issues in human biometeorology. To quantify these effects, the UTCI is developed in Cooperation in Scientific and Technical Research (COST) action 730. The main aim of the index is to present an environmental-physiological evaluation model to increase the plans related to health and welfare in public climate service, public health systems, prevention designs, and climate effects research (Broede et al. 2012).

In order to calculate UTCI, the data of average wind speed, relative humidity, and sunshine duration and mean radiant temperature (MRT) were used. Sunshine duration was used to estimate direct solar radiation (Rs) (which is required as input to the mean radiant temperature) and relative humidity was used to estimate water vapor pressure based on daily maximum and minimum temperature values (Lemke & Kjellstrom 2012). MRTs were calculated from air temperature, global temperature, and wind speed.

To calculate the global temperature, first we need to have Rs. If Rs is unknown, it can be calculated with the angstrom formula (1):
formula
(1)
where no actual solar radiation data are available and no calibration has been carried out, for improved as and bs parameters, the values of 0.25 and 0.50 are recommended, respectively, by the Food and Agricultural Organization of the United Nations (Allen et al. 1998). Here n is actual duration of sunshine [hour] and N is maximum possible duration of sunshine of daylight hours [hour]. Maximum possible duration of sunshine, N, and the sunset hour angle, , are calculated from Equations (2) and (3), respectively:
formula
(2)
formula
(3)
The latitude, expressed in radians is considered positive for the northern hemisphere and negative for the southern hemisphere. The conversion from decimal degree to radians and the solar declination, , are given by:
formula
(4)
formula
(5)
The extraterrestrial radiation, Ra, for each day of the year and for different latitudes can be estimated from the solar constant, the solar declination, and the time of the year by:
formula
(6)
where Gsc solar constant equals to 0.0820 MJ m−2 min−1, and the inverse relative Earth-Sun, dr, is calculated from Equation (7):
formula
(7)
Mean radiant temperature (MRT) calculated from Equation (8) is:
formula
(8)
For converting the relative humidity to vapor pressure carried out based on Brice & Hall (2013) (Equation (9)):
formula
(9)
Use this empirically derived formula for GT (Equation (10)):
formula
(10)

Finally, UTCI is computed by the regression equation found at www.UTCI.org. The rating categories of the UTCI are shown in Table 3.

Table 3

UTCI equivalent temperature and stress category (Pappenberger et al. 2015)

UTCI (°C) rangeStress category
>46 Extreme heat stress 
38–46 Very strong heat stress 
32–38 Strong heat stress 
26–32 Moderate heat stress 
9–26 No thermal stress 
0–9 Slight cold stress 
0 to −13 Moderate cold stress 
−13 to −27 Strong cold stress 
−27 to −40 Very strong cold stress 
<− 40 Extreme cold stress 
UTCI (°C) rangeStress category
>46 Extreme heat stress 
38–46 Very strong heat stress 
32–38 Strong heat stress 
26–32 Moderate heat stress 
9–26 No thermal stress 
0–9 Slight cold stress 
0 to −13 Moderate cold stress 
−13 to −27 Strong cold stress 
−27 to −40 Very strong cold stress 
<− 40 Extreme cold stress 
The tourism climatic index (TCI) represents a quantitative evaluation of world climate for tourism purposes and is an absolute measure of the climatic well-being of tourists. The TCI is aimed at tourists involved in sight-seeing or light outdoor activities. The TCI was originally defined by Mieczkowski (1985) as a weighted sum of several factors. TCI is calculated according to Equation (11):
formula
(11)
where CID is a daytime comfort index, consisting of the mean maximum air temperature (°C) and the mean minimum relative humidity RH (%), CIA is the daily comfort index, consisting of the mean air temperature (°C) and the mean relative humidity (%), R is the precipitation (mm), S is the daily sunshine duration (h), and W is the mean wind speed (m/s). Due to the weighting factor (a value for TCI of 100), every factor can reach 5 points. The value of the TCI varies between 100 (ideal) to <10 (impossible). The rating categories of TCI are shown in Table 4.
Table 4

Classification scheme for the tourism climatic index (Mieczkowski 1985)

TCI rangeDescription
90–100 Ideal 
80–89 Excellent 
70–79 Very good 
60–69 Good 
50–59 Acceptable 
40–49 Marginal 
30–39 Unfavorable 
20–29 Very unfavorable 
10–19 Extremely unfavorable 
<10 Impossible 
TCI rangeDescription
90–100 Ideal 
80–89 Excellent 
70–79 Very good 
60–69 Good 
50–59 Acceptable 
40–49 Marginal 
30–39 Unfavorable 
20–29 Very unfavorable 
10–19 Extremely unfavorable 
<10 Impossible 

Mann–Kendall test

In order to study the trend in time series, the Mann–Kendall test is used assuming the rejection of ‘lack of trend in time series’. The nonparametric Mann–Kendall test was developed by Mann (1945) and Kendall (1962) based on data ranking in time series not following a certain statistical distribution (Partal & Kahya 2006; Modarres & da Silva 2007; Yue et al. 2002). The test is suitable for data that do not follow a normal distribution, and supports multiple observations per time series (Kampata et al. 2008). The Mann–Kendall test is given as:
formula
(12)
formula
(13)
formula
(14)
where n is the number of data points, t is the number of ties for i value, and m is the number of tied values. Then, Equations (12) and (13) were used to compute the test statistic Z from the following equation:
formula
(15)

A positive value of Z indicates an increasing trend and a negative value indicates a decreasing trend. The null hypothesis, H, that there were no trends in the records is either accepted or rejected depending on whether the computed Z statistic is less than or more than the critical value of Z statistics obtained from the normal distribution table at the 5% significance level (Kampata et al. 2008).

In this paper the trend of compound index changes is computed based on decade. The trend of compound index (yj) can be computed using the least square linear regression equation:
formula
(16)
where yj is the value of index in year j and ej is a disturbance term (residual) with mean zero. The regression coefficient b gives the change per decade (Kanji 2006).

RESULTS AND DISCUSSION

Common index of temperature and precipitation

Most extreme climate events result from combined climate variables. For instance, high temperature with low rainfall may cause heat waves and drought. The changes in threshold ranges for four precipitations (CD, CW, WD, and WW) at Ramsar meteorological station during 1981–2015 show that the possibility of occurrence is different and dry modes have the highest possibilities since 1998 (Figure 2). There is a clear correspondence between temperature and precipitation compound index. There are long-term changes in temperature and rainfall annual series based on field observations in Figures 37.

Figure 2

Threshold exceedance of the four joint temperature and precipitation modes (CD, WD, CW, WW) for Ramsar station for the period 1981–2015 [days per years].

Figure 2

Threshold exceedance of the four joint temperature and precipitation modes (CD, WD, CW, WW) for Ramsar station for the period 1981–2015 [days per years].

Figure 3

Decadal trends in the occurrences of the cold/dry extremes (CD) (left) and cool/wet extremes (CW) (right), for the period 1981–2015 (days/decade).

Figure 3

Decadal trends in the occurrences of the cold/dry extremes (CD) (left) and cool/wet extremes (CW) (right), for the period 1981–2015 (days/decade).

Figure 4

Time series of the cold/dry (CD) extremes for Abadan (right) and Tabriz (left) stations for the period 1981–2015 (days/decade).The solid (dashed) black lines indicate trend lines, and the equations given in the top of each diagram represent specifications of the trends.

Figure 4

Time series of the cold/dry (CD) extremes for Abadan (right) and Tabriz (left) stations for the period 1981–2015 (days/decade).The solid (dashed) black lines indicate trend lines, and the equations given in the top of each diagram represent specifications of the trends.

Figure 5

Time series of the cool/wet (CW) extremes for Hamedan (left) and Sanandaj (right) stations for the period 1981–2015 (days/decade). The solid (dashed) black lines indicate trend lines, and the equations given in the top of each diagram represent specifications of the trends.

Figure 5

Time series of the cool/wet (CW) extremes for Hamedan (left) and Sanandaj (right) stations for the period 1981–2015 (days/decade). The solid (dashed) black lines indicate trend lines, and the equations given in the top of each diagram represent specifications of the trends.

Figure 6

Decadal trends in the occurrences of the warm/dry (WD) (left) and the warm/wet (WW) (right) extremes for the period 1981–2015 (days/decade).

Figure 6

Decadal trends in the occurrences of the warm/dry (WD) (left) and the warm/wet (WW) (right) extremes for the period 1981–2015 (days/decade).

Figure 7

Time series of the warm/dry (WD) extremes for Kish Island (left) and Chabahar (right) stations for the period 1981–2015 (days/decade).

Figure 7

Time series of the warm/dry (WD) extremes for Kish Island (left) and Chabahar (right) stations for the period 1981–2015 (days/decade).

Cold/dry days (CD) and cold/wet days (CW)

The annual trends of CD compound index based on temperature and precipitation data in Iran are presented in Figure 3(left). This index has a decreasing trend in most stations. The rate of change in the study period based on location between 1 and 10 days in each decade and 1 and 3 days in each decade is decreasing. Except for 15 cases with significant negative trends (p < 0.05) that are mainly located in the northwest, we cannot find significant negative trends for the annual CD over the whole country, but the decreasing trends are dominant. The changes depend on the location over the course of 1981–2015 in CD mode, and decrease between a large spread of 1–10 days per decade, to a much tighter spread of 1–3 days per decade.

The largest and the smallest significant negative trends of CD days occurred in Tabriz station, about (−) 6.5 day per decade and in Abadan station about (−) 3.5 day per decade (Figure 4). Negative trends in CD were larger than those in CW at most stations. The positive trend in this index was observed in some dispersed small areas (Shahrekord, Sabzevar, Gorgan, Rasht, and Torbat e-heydareyeh stations). The magnitudes of the increasing trends in annual CD ranged between (+) 3 days per decade at Shahrekord station and (+) 0.3 days per decade at Sabzevar station.

In some regions, such as Khoramabad and Bojnourd, there are very weak and indiscernible trends. Also, coastal areas of the Caspian Sea and coastal strips of the Persian Gulf clearly show higher annual CD days.

Distinguishable negative trends of annual CD were found to be in good agreement with the results obtained for other territories in Europe (Beniston 2009), where a tendency towards colder years has been identified.

The CW index is defined as the number of days where T25/P75. Here, T25 indicates temperature occurrences below the 25% quantile, while P75 denotes occurrences above the 75% quantile. The outputs of the statistical tests in the annual CW series over the period of 1981–2015 are shown in Figure 3 (right). Except in parts of the south-west country in the south of the Zagros Mountain range and some dispersed stations, other parts of the country showed negative trends for the CW index. However, the MK test indicated statistically significant trends at 0.05 levels for only the two stations of Ramsar and Sannandaj in this index. In other words, the number of cold/wet days at the eight stations (17%) had an increase over the study period while the other stations (83%) showed a decreasing trend.

The largest negative trends of the CW index can be seen in the northern half of the country and the southern part, especially the south-east of the country. The largest negative trends in CW have been found in Sanandaj station, showing a decrease of approximately (−) 1.68 days per decade and Hamedan (Airport) station with the magnitude of approximately (−) 1.12 days per decade (Figure 5).

In some stations, there are very weak positive trends and fluctuations such that there is no significance. In other words, there are no significant trends and regular dispersal of CW in these areas.

Generally, Iran experienced a negative trend for CD and CW indices. It is reasonable to assume that CD and CW extreme indices are related to general circulation patterns over Iran. The results are consistent with the results of the study by Hao et al. (2013) for the Middle East region.

Warm/dry days (WD) and warm/wet days (WW)

Based on our results, the index trend of warm/dry extremes has increased all over the country. The results of the WD index (80% of the stations) show the increase at more than 80% of stations (Figure 6(left)). Significant positive trends of WD for mountainous ranges, dry and semi-dry climatic regions were recognized. In addition, the significant increasing trends in annual WD varied between (+) 3.7 days per decade at Chabahar station and (+) 14.5 days per decade at the Kish Island station (Figure 7). In recent years, in many regions such as central Africa, eastern Australia, and some parts of Russia, the WD has increased too (Hao et al. 2013). Except for Shahrekord station located in the south-west of the country, the other areas of the country experienced a positive trend in WD. This may be due to the cooling trends found for the southern half of the Zagros Mountains and Shahrekord station (Parak et al. 2015). Similar behavior in fluctuations in observational extreme precipitation and temperature data sets for 1961–2010 can also be seen in Parak et al. (2015). To answer this anomaly needs more research and more precise demands. It should be noted that in some regions, such as southern Zagros (west) and eastern Alborz (north-east), positive trends were non-significant for the number of the WD index.

Figure 7 depicts the time series of warm/dry (WD) for Kish Island and Chabahar stations. The trend line shows an increase in the annual WD since 1981 for Kish Island and Chabahar stations. These stations are located in coastal and marine climates and extreme conditions were modified by marine conditions. Although in these regions temperatures have been affected by the general trend of global warming over the country, local trends can be due to urbanization in these areas. This pattern is important in occurrences of weather phenomena like heatwaves and droughts that cause social and environmental damage (Hao et al. 2013). In addition, factors such as local and regional air pollution may influence the nature and magnitude of the climatic trends in different ways (De Sario et al. 2013).

Positive trends for the WD were found across most of the country which is in parallel with IPCC findings. Generally, the period from 1983 to 2012 is much warmer than any 30-year period in the last 1,400 years in the northern hemisphere.

The warm/wet (WW) combination trends are shown in Figure 6 (left). The results of linear regression indicate that about 42% of the studied stations show a positive trend in extreme warm/wet index during the past 35 years. However, at the 95% confidence level, few of the observed incremental trends are significant and are mostly located in the north and north-west of the country. Based on the above, most positive trends are relatively weak and non-significant; however, the strongest is at Noushahr station (north of the country) with a trend of (+) 0.56 per decade.

The result also indicated no trends of warm/wet in coastal strips of the Persian Gulf and the south-west and some scattered areas in the country. If the country is divided into two parts with an artificial northeast–southwest line, most of the positive trends are seen on the north side of the line (Figure 6 (right)). Generally, the warm/wet (WW) index has incremental trends over the high-latitude regions, whereas the warm/dry indicator has increased in most areas. Our result is consistent with the trend observed in different countries and at the global scale (Beniston 2009; Hao et al. 2013).

Although there is a similar behavior between WW and WD, magnitude and spatial coherence of the earlier index is very weak. Our results indicate that the stations located in the Caspian Sea region such as Ramsar, Noushahr, Rasht have experienced warm/wet days in most years of the period, whereas other stations have experienced some years with no warm/wet days. In the stations located in the south of the country, the WW index has been zero in most years. In the warm months, especially in summer, the anticyclonic circulation is strengthened in the west of the Arabian Peninsula, and its extension over this area blocks moisture advection into the south of Iran from the moisture source (Parak et al. 2015). This anticyclonic circulation prevents the air from rising and causing precipitation in the area. Such patterns may lead to differences in local climate that may be affected by global warming which tends to make wet areas wetter and dry areas drier (Romm 2015). But the effects differ in different parts of the country. The comparison of results for the indices of WD and WW days showed a similar behavior between the two indices, but the magnitude and spatial consistency of WW days were much less than WD days.

UTCI

The results of this study indicate that most of the country generally experienced a reduction in the number of days with cold stress. On the other hand, the increase in terms of warm stress also affects climate comfort conditions in the country. In other words, decreasing trends were found in the annual occurrence of the no thermal stress class, at 45 stations (95%), which was statistically significant (p < 0.05) (Figure 8(left)). Negative significant trend rates in UTCI in the whole of the country lay in the range of (−) 4 days per decade (Ahvaz station) to −20 days per decade at (Gorgan station).

Figure 8

Decadal trends in occurrences of the UTCI in the range of 9–26 °C (left) and in the range of 32–38 °C (right) for the period 1981–2015 (days/decade).

Figure 8

Decadal trends in occurrences of the UTCI in the range of 9–26 °C (left) and in the range of 32–38 °C (right) for the period 1981–2015 (days/decade).

In a region like the Zagros Mountain range, the negative trends obtained are not so strong. From the results, it is clear that the decreasing trends of the no thermal stress class in the north, northwest, and southeast regions of Iran were stronger than those in the other areas during the last decades.

Our results revealed that positive trends of UTCI were only experienced at Ardebil and Shiraz stations in the range of 9–26 °C. The significant positive trend for this index was found in the western regions of the Caspian Sea at Ardebil station by (+) 6.9 days per decade (Figure 9).

Figure 9

Time series of the events in the range of 9–26 °C extremes for Gorgan (left) and Ahvaz (right) stations for the period 1981–2015 (days/decade).

Figure 9

Time series of the events in the range of 9–26 °C extremes for Gorgan (left) and Ahvaz (right) stations for the period 1981–2015 (days/decade).

Based on the results presented in Figure 8 (left), annual occurrence of strong heat stress (32–38 °C) increased approximately all over the country. Furthermore, the value of the significant increasing trend in this index varied from (+) 0.5 to (+) 17.3 days per decade at Zanjan and Chabahar stations, respectively. Less than one-fifth of the country, including regions in the south and southwest, experienced negative trends or no trends in this index, but stations such as Ahvaz, Abadan, Boushehr, Fasa, Kish Island, Bandar e-lengeh, and Bandarabas showed an increasing trend in UTCI events in the annual occurrence of the very strong heat stress (38–46 °C) class.

Generally, compared with other compound quantities of temperature and precipitation index, UTCI is consistent with warming and increases of the warm modes' frequency and the probability of the intensity of the thermal stress is not diminished and has shifted to moderate and high levels. Global warming raises the average temperature which causes heat waves to become more intense and more frequent. For the same reason, heat waves will last longer and cover a large region. In recent years, the severity and frequency of drought has increased due to global warming in many parts of the world (Romm 2015). The results show factors such as topography and latitude play an important role in UTCI distribution. In other words, the annual UTCI increases with decreasing latitude throughout most of Iran.

TCI

Trends in TCI, including the number of days with TCI ≥ 60, and the number of days with TCI ≥ 80 are shown in Figure 10. The results indicated that the majority of the trends in the annual TCI60 and TCI80 series were positive. The majority of the country showed a positive significant trend in TCI60. Significant positive trend rates in the TCI60 series were in the range of (+) 1.6 days per decade at Rasht station to (+) 16 days per decade at Ardebil station.

Figure 10

Decadal trends in the occurrences of the number of days with TCI ≥ 60 (left) and the number of days with TCI ≥ 80 (right) for the period 1981–2015 (days/decade).

Figure 10

Decadal trends in the occurrences of the number of days with TCI ≥ 60 (left) and the number of days with TCI ≥ 80 (right) for the period 1981–2015 (days/decade).

The greatest values of this index can be observed in the provinces located in the northwest of the country.

On the contrary, the magnitude of significant negative trends at Chabahar and Gorgan stations located in the southeast and north of the country, respectively, showed a significant negative trend. The steepest negative trend was seen at Chabahar station (southeast), showing a decrease of (−) 8 days per decade.

Comparison between TCI80 and TCI60 indicated that the TCI80 index has a more irregular distribution than positive and negative trends in the country. Over 75% of the stations studied in the country showed an increase in the annual occurrence of TCI80 and 17% of these have significant increasing trends (Figure 10(right)). The largest positive significant trend in TCI80 was found at Ardebil station, which shows an increase of 10 days per decade.

About 20% of the stations, mostly in the northwest and southeast of the country, had negative trends. The largest negative trend occurred at Iranshahr station, showing a decrease of (−) 2.2 days per decade. Generally, the relationships that have been observed may be explained by several physically different mechanisms. Such correlations may occur because of changes in cloud cover or changes in precipitation and the consequent changes in the heat balance of the regions under consideration (Zhao & Khalil 1993). Thus, it is worth mentioning the issue of whether the observed trends were related to a long-term process like global warming or if they were part of a multi-decadal natural oscillation. Further studies would be interesting to compare the trends found in our study with the results of other tests to detect trends.

CONCLUSION

This paper has attempted to provide an appropriate consideration of extreme climate process by using compound extreme index with a primary focus on extreme events. There were long-term records for a few stations before 1981. There are some limitations for most stations, such as inhomogeneities and uncertainties in their data sets caused by changes like location, exposure, instrumentation, and observation practice, and also from missing data. We used synoptic stations whose data were reliable, continuous, and long. These stations cover most of the climate types and are evenly geographically distributed throughout the country. These indices were calculated and analyzed for 47 sites during 1981–2015 to provide a general overview of climate change in Iran. Climatic variables such as temperature, precipitation, humidity, sunshine, and wind speed were analyzed. The results of non-parametric tests highlighted the statistically significant spatially coherent trends in compound values of temperature and precipitation index (using the 25% and 75% value levels) corresponding to a warming trend in the country. It was concluded that in more than 80% of Iran the frequency of the warm modes has increased while the frequency of cold modes has decreased but with smaller magnitudes. Generally, Iran experienced a negative trend for the CD and CW index and a positive trend in CW was observed in some dispersed small areas.

More than 97% of the stations exhibited a positive trend for the annual WD index. Positive trends for the WD were found all over the country which is in parallel with IPCC findings. The results of linear regression indicate that about 42% of the studied stations show a positive trend in extreme warm/wet index that are seen in the northern half of the country. The rate of the significant increasing trends in annual WD varied from (+) 3.7 days per decade at Chabahar station to (+) 14.5 days per decade at Kish Island station. Based on the results of the analysis, apart from a few stations and, more specifically, in Shahrekord station, the overall patterns of compound values of temperature and precipitation show the same trend, and responding to these differences definitely requires further investigation. Both WD and WW indices have similar behavior, but the magnitude and spatial consistency of WW days were much less than WD days. The results show that most of the country generally experienced a reduction in the number of days with cold stress and an increase in warm stress. These factors affect climate comfort conditions in the country.

Based on the results, UTCI is consistent with warming and increases of the warm mode frequency and the probability of the intensity of the thermal stress is not diminished and has shifted to moderate and high levels. Global warming raises the average temperature that causes heatwaves to become more intense and more frequent. At the same time, the occurrence of TCI and UTCI in strong heat stress (32–38 °C) events has increased over most of the country, consistent with increases in the warm modes' frequency. The majority of the country showed a positive significant trend in TCI60, and TCI80 index has a more irregular distribution than positive and negative trends in the country. The distribution of the annual compound index trends indicated that the negative and positive significant trends mainly occurred in the northwest of Iran. The results also suggest the need for further investigation on the intervention of local factors in the environment, which could be one of the major causes of climate change. The findings of this study confirmed the results of previous studies (e.g., IPCC 2007). Although our results are consistent with the findings of other studies such as a Middle East regional study (Zhang et al. 2005, 2011) and trends of climate extremes over Iran (Parak et al. 2015), the results also suggest the need for further investigation by using gridded daily data and remote sensing data on local anthropogenic factors, which could be one of the major causes of climate change in Iran.

ACKNOWLEDGEMENTS

The authors would like to thank the anonymous reviewers for their comments and suggestions which have contributed to improving the manuscript. We also thank the Islamic Republic of Iran Meteorological Organization (IRIMO) for providing the data.

REFERENCES

AghaKouchak
A.
Cheng
L.
Mazdiyasni
O.
Farahmand
A.
2014
Global warming and changes in risk of concurrent climate extremes: insights from the 2014 California drought
.
Geophysical Research Letters
41
,
8847
8852
.
doi:10.1002/2014GL062308
.
Alexander
L. V.
Zhang
X.
Peterson
T. C.
Caesar
J.
Gleason
B.
Klein Tank
A. M. G.
Haylock
M.
Collins
D.
Trewin
B.
Rahim
F.
Tagipour
A.
Kumar Kolli
R.
Revadekar
J. V.
Griffiths
G.
Vincent
L.
Stephenson
D. B.
Burn
J.
Aguilar
E.
Brunet
M.
Taylor
M.
New
M.
Zhai
P.
Rusticucci
M.
Vazquez Aguirre
J. L.
2006
Global observed changes in daily climate extremes of temperature and precipitation
.
Journal of Geophysical Research Atmospheres
111
,
D05109
.
doi:10.1029/2005JD006290
.
Allen
G. R.
Pereira
S. P.
Raes
D.
Smith
M.
1998
Crop Evapotranspiration – Guideline for Computing Crop Water Requirements
.
FAO – Food and Agriculture Organization of the United Nations
,
Rome
,
Italy
.
Arsenovic
P.
Tosic
I.
Unkasevic
M.
2013
Seasonal analysis of warm days in Belgrade and Nis
.
Journal of the Geographical Institute Jovan Cvijic, SASA
63
(
4
),
1
10
.
doi:10.2298/IJGI1304001A
.
Beniston
M.
2009
Trends in joint quantiles of temperature and precipitation in Europe since 1901 and projected for 2100
.
Geophysical Research letters
36
,
7
.
https://doi.org/10.1029/2008GL037119
.
Brice
T.
Hall
T.
2013
The weather calculator
.
Available from: www.srh.noaa.gov/epz/?n=wxcalc.] [Accessed 13 September 2016.]
Broede
P.
Fiala
D.
Blazejczyk
K.
Holmer
I.
Jendritzky
G.
Kampmann
B.
Tinz
B.
Havenith
G.
2012
Deriving the operational procedure for the Universal Thermal Climate Index (UTCI)
.
International Journal of Biometeorology
56
(
3
),
481
494
.
doi:10.1007/s00484-011-0454-1
.
Climate Change Science Program (CCSP)
2008
In:
Weather and Climate Extremes in A Changing Climate. Regions of Focus: North America, Hawaii, Caribbean, and U.S. Pacific Islands
.
A Report by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research
(
Karl
T. R.
Meehl
G. A.
Miller
C. D.
Hassol
S. J.
Waple
A. M.
Murray
W. L.
, eds).
Department of Commerce, NOAA's National Climatic Data Center
,
Washington DC
,
USA
, p.
164
.
De Sario
M.
Katsouyanni
K.
Michelozzi
P.
2013
Climate change, extreme weather events, air pollution and respiratory health in Europe
.
European Respiratory Journal
42
,
826
843
.
doi:10.1183/09031936.00074712
.
Fan
L.
Chen
D.
2016
Trends in extreme precipitation indices across China detected using quantile regression
.
Atmospheric Science Letters
17
,
400
406
.
doi/full/10.1002/asl.671
.
Filho
W. L.
Musa
H.
Cavan
G.
O'Hare
P.
Seixas
J.
2016
Climate Change Adaptation, Resilience and Hazards
.
Springer
.
doi:10.1007/978-3-319-39880-8
.
Hao
Z.
AghaKouchak
A.
Phillips
T. J.
2013
Changes in concurrent monthly precipitation and temperature extremes
.
Environmental Research Letters
8
,
034014
.
doi:10.1088/17489326/8/3/034014
.
IPCC
2007
In:
Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change
(
Solomon
S.
Qin
D.
Manning
M.
Chen
Z.
Marquis
M.
Averyt
K. B.
Tignor
M.
Miller
H. L.
, eds).
Cambridge University Press
,
Cambridge
,
UK
and New York, NY, USA
, p.
966
.
IPCC
2013
In:
Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
(
Stocker
T. F.
Qin
D.
Plattner
G. K.
Tignor
M. M. B.
Allen
S. K.
Boschung
J.
Nauels
A.
Xia
Y.
Bex
V.
Midgley
P. M.
, eds).
Cambridge University Press
,
Cambridge
,
UK
and New York, NY, USA
, p.
966
.
Kampata
J. M.
Parida
B. P.
Moalafhi
D. B.
2008
Trend analysis of rainfall in the headstreams of the Zambezi River Basin in Zambia
.
Physics and Chemistry of the Earth, Parts A/B/C
33
,
621
625
.
Kanji
G. K.
2006
100 Statistical Tests
.
SAGE Publications Ltd
,
ISBN-13 978 1 4129 2376 7 (Pbk). Available from: https://www.amazon.com/100-Statistical-Tests-Gopal-Kanji/dp/141292376X.
Kendall
M. G.
1962
Rank correlation methods
, 3rd edn.
Hafner, New York, USA
.
Klok
E. J.
Klein Tank
A. M. G.
2008
Updated and extended European dataset of daily climate observations
.
International Journal of Climatology
29
,
1182
1191
.
doi:10.1002/joc.1779
.
Leonard
M.
Westra
S.
Phatak
A.
Lambert
M.
Hurk
B. V. D.
McInnes
K.
Risbey
J.
Schuster
S.
Jakob
D.
Stafford-Smith
M.
2014
A compound event framework for understanding extreme impacts
.
WIREs Climate Change
5
,
113
128
.
doi:10.1002/wcc.252
.
Liu
C.
Allan
R. P.
Huffman
G. J.
2012
Co-variation of temperature and precipitation in CMIP5 models and satellite observations
.
Geophysical Research Letters
39
,
L13803
.
Lopez-Moreno
J. I.
Vicente-Serrano
S. M.
Moran-Tejeda
E.
Lorenzo-Lacruz
J.
Kenawy
A.
Beniston
M.
2011
Effects of the North Atlantic Oscillation (NAO) on combined temperature and precipitation winter modes in the Mediterranean mountains: observed relationships and projections for the 21st century
.
Global and Planetary Change
77
,
62
76
.
doi:10.1016/j.gloplacha.2011.03.003
.
Mann
H. B.
1945
Nonparametric tests against trend
.
Econometrica
13
,
245
259
.
DOI: 10.2307/1907187
.
Miao
C.
Sun
Q.
Duan
Q.
Wang
Y.
2016
Joint analysis of changes in temperature and precipitation on the Loess Plateau during the period 1961–2011
.
Climate Dynamics
47
,
3221
.
doi:10.1007/s00382-016-3022-x
.
Mieczkowski
Z.
1985
The Tourism Climatic Index: a method of evaluating world climate for tourism
.
The Canadian Geographer
29
,
220
233
.
doi/abs/10.1111/j.1541-0064.1985.tb00365.x
.
Moazenzadeh
R.
Mohammadi
B.
Shamshirband
S.
Chau
K.
2018
Coupling a firefly algorithm with support vector regression to predict evaporation in northern Iran
.
Engineering Applications of Computational Fluid Mechanics
12
(
1
),
584
597
.
doi:10.1080/19942060.2018.1482476
.
Modarres
R.
da Silva
V. P. R.
2007
Rainfall trends in arid and semi-arid regions of Iran
.
Journal of Arid Environments
70
,
344
355
.
Pappenberger
P.
Jendritzky
G.
Staiger
H.
Dutra
E.
Di Giuseppe
F.
Richardson
D. S.
Cloke
H. L.
2015
Global forecasting of thermal health hazards: the skill of probabilistic predictions of the Universal Thermal Climate Index (UTCI)
.
International Journal of Biometeorology
59
,
311
323
.
doi:10.1007/s00484-014-0843-3
.
Parak
F.
Roshani
A.
BodaghJamali
J.
2015
Trends and anomalies in daily climate extremes over Iran during 1961–2010
.
Journal of Environmental and Agricultural Sciences
.
2
,
11
.
ISSN: 2313–8629. Available from: http://www.agropublishers.com/files/Article%2011.pdf
Partal
T.
Kahya
E.
2006
Trend analysis in Turkish precipitation data
.
Hydrological Processes
20
,
2011
2026
.
Project team ECA&D, Royal Netherlands Meteorological Institute KNMI
2013
EUMETNET/ECSN optional programme: European Climate Assessment & Dataset (ECA&D) Algorithm Theoretical Basis Document (ATBD), Version: 10.7
.
Rahimzadeh
F.
Asgari
A.
Fattahi
E.
2009
Variability of extreme temperature and precipitation in Iran during recent decades
.
International Journal of Climatology
29
,
329
343
.
https://doi.org/10.1002/joc.1739
.
Romm
J.
2015
Climate Change (What Everyone Needs To Know)
.
Oxford University Press
,
New York
,
USA
.
Tabari
H.
ShiftehSomee
B.
RezaeianZadeh
M.
2011
Testing for long-term trends in climatic variables in Iran
.
Atmospheric Research
100
,
132
140
.
doi:10.1016/j.atmosres.2011.01.005
.
Tayanç
M.
Dalfes
H. N.
Karaca
M.
Yenigün
O.
1998
A comparative assessment of different methods for detecting inhomogeneities in Turkish temperature data set
.
International Journal of Climatology
18
,
561
578
.
doi:10.1002/(sici)1097-0088(199804)18:5 < 561::aid-joc249 > 3.0.co;2-y
.
Tsidu
G. M.
2012
High-resolution monthly rainfall database for Ethiopia: homogenization, reconstruction, and gridding
.
Journal of Climate
25
,
8422
8443
.
doi:10.1175/jcli-d-12-00027.1
.
Vincent
L. A.
Wang
X. L.
Milewska
E. J.
Wan
H.
Feng
Y.
Swail
V.
2012
A second generation of homogenized Canadian monthly surface air temperature for climate trend analysis
.
Journal of Geophysical Research Atmospheres
117
,
D18110
.
Wang
X. L. L.
2008
Accounting for autocorrelation in detecting mean-shifts in climate data series using the penalized maximal t or F test
.
Journal of Applied Meteorology and Climatology
47
,
2423
2444
.
doi/10.1175/2008JAMC1741.1
.
Wang
X. L.
Feng
Y.
2013
RH testsV4 User Manual
.
Climate Research Division, Atmospheric Science and Technology Directorate, Science and Technology Branch, Environment Canada
, p.
28
.
Wang
X. L.
Wen
Q.
Wu
Y.
2007
Penalized maximal t test for detecting undocumented mean change in climate data series
.
Journal of Applied Meteorology and Climatology
46
,
916
931
.
doi:10.1175/JAM2504.1
.
Wijngaard
J.
Tank
A. K.
Können
G.
2003
Homogeneity of 20th century European daily temperature and precipitation series
.
International Journal of Climatology
23
,
679
692
.
doi:10.1002/joc.906
.
World Meteorological Organization
2009
Guidelines on Analysis of Extremes in A Changing Climate in Support of Informed Decisions for Adaptation. Climate Data and Monitoring WCDMP-No. 72
.
WMO
,
Geneva
,
Switzerland
.
World Meteorological Organization
2011
Guide to Climatological Practices. WMO-No. 100
.
WMO
,
Geneva
,
Switzerland
.
Yozgatligil
C.
Yazici
C.
2015
Comparison of homogeneity tests for temperature using a simulation study
.
International Journal of Climatology
36
,
62
81
.
doi:10.1002/joc.4329
.
Zhang
X.
Aguilar
E.
Sensoy
S.
Melkonyan
H.
Tagiyeva
U.
Ahmed
N.
Kutaladze
N.
Rahimzadeh
F.
Taghipour
A.
Hantosh
T. H.
Albert
P.
Semawi
M.
Karam Ali
M.
Halal Said Al-Shabibi
M.
Al-Oulan
Z.
Zatari
T.
Al Dean Khelet
I.
Hammoud
S.
Demircan
M.
Eken
M.
Adiguzel
M.
Alexander
L.
Peterson
T.
Wallis
T.
2005
Trends in middle east climate extremes indices during 1930–2003
.
Journal of Geophysical Research
110
(
D22
),
104
.
doi:10.1029/2005JD006181
.
Zhang
X.
Alexander
L.
Hegerl
G. C.
Jones
P.
Tank
A. K.
Peterson
T. C.
Trewin
B.
Zwiers
F. W.
2011
Indices for monitoring changes in extremes based on daily temperature and precipitation data
.
WIREs Climate Change
2
,
851
870
.
doi:10.1002/wcc.147
.
Zhao
W.
Khalil
M. A.
1993
The relationship between precipitation and temperature over the contiguous United States
.
J. Climate
6
,
1232
1236
.
https://doi.org/10.1175/1520-0442
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).