An examination of the trends in sunshine hours over Iran



For the purpose of assessing solar energy resources in different parts of Iran, this article provides a synopsis of the spatial and temporal variability of sunshine hours at 37 first-order synoptic stations during the period 1981–2007. Annual and seasonal variations of sunshine duration were determined for four distinct regions within the country. By averaging the time series of sunshine hours in each subregion and standardizing them, four regional representative time series were developed. The results indicate, over all regions, the annual pattern of sunshine duration exhibited large sinusoidal increases and decreases, with minima occurring in 1990 and 2003. The trends of sunshine duration indicated that the sign of the seasonal and annual trends for the vast majority stations has been positive. The maximum positive trend was found across the western parts of the country for all seasons. The spring was found to have a negative trend in sunshine hours at only 2 of the 37 stations. On the annual scale, the change rates in sunshine hours from western to eastern part of Iran have gradually declined over time. The highest positive annual trend was found at Sanandaj station on the west side of Zagross Mountain, with of rate 253 h per decade. Given the increasing trends observed at many stations in Iran over the last 25 years, and recognizing the natural high sunshine duration experienced across the region, a strong case can be made for the introduction of solar energy across the country. Copyright © 2012 Royal Meteorological Society

1. Introduction

Sunshine is one of the most important ingredients to life on Earth. Over the course of the year, the Sun's rays warm the land and oceans and are a direct factor in photosynthesis, driving the hydrologic cycle, and ultimately are the primary driver of the Earth's weather and climate. Sunshine is a combination of light and radiant heat emitted by the Sun. Furthermore, it is related directly to solar energy resources that are very important as a mean of mitigation of climate change to meet goals for reduction of greenhouse gas emissions.

Ground-based observations of sunshine duration are typically recorded using a pyranometer or a pyrheliometer, both of which measure direct solar irradiance by converting heat into an electrical signal and determining the sunshine duration. Surveys on the changes of sunshine duration can be useful both for changes of this essential climatic factor and also as proxy to explore changes in shortwave forcing at the Earth's surface, in particular for where accurate irradiances are not measured or recorded. It is essential to understand the potential of the solar energy resource in different regions of the country due to changes of major climate variables including cloud cover, duration of sunshine hours and solar flux at the Earth's surface.

Several articles have explored the relationship between sunshine duration and climate change. A significant negative trend for annual sunshine hours at over 70% of the selected synoptic stations was found in Turkey (Aksoy, 1999). In China, the annual number of sunshine hours has decreased over the period 1961–1998 and 1965–1999 (Kaiser and Qian, 2002; Chen et al., 2006; Yang et al. 2009). In the United States, Stanhill and Cohen (2005) showed during the 20th Century regionally, in the northwest quarter of the country, sunshine duration increased at nine sites and decreased at three. In the three other quarters of the U.S. the numbers of sites with increasing and decreasing trends were equal. After 1950, a larger proportion of series showed decreases in sunshine duration, and more sites showing decreasing trends were found in the northeast and in west and south of USA, but these regional differences were not statistically significant. Sanchez-Lorenzo et al. (2008) found a direct correlation between the dimming and brightening phenomena of the Sun with a negative trend in sunshine duration for Western Europe. In addition, they found that the spring and annual series exhibited negative sunshine duration trends, while a positive trend was apparent in the winter for this area (Sanchez-Lorenzo et al., 2008). A similar result was found in Tibet for the period 1971–2005, where seasonal decreases in sunshine duration were evident in summer while increases were observed in winter (Liu et al., 2002; Du et al., 2007).

There is a number of research articles related to sunshine duration over Iran, particularly in quantifying estimates in lieu of actual solar measurements. The entire data base of the Iranian meteorological stations was used to develop a simple relationship between the sunshine duration and cloudiness records (Samimi, 1994; Kamali and Moradi, 2004). Sabziparvar and Shetaee (2007) developed a general height-dependent formula for the prediction of the direct and diffuse monthly average daily solar radiation for 64 mountainous arid and semi-arid locations outspread over the country. Saffaripour and Mehrabian (2009) have estimated the solar radiation intensity for Yazd province in the centre of Iran. They used seven geographical and climatological variables including the maximum possible radiation in the absence of the atmosphere and sunshine hours measured at meteorological synoptic stations. They found the solar declination angle could adequately predict the solar radiation with satisfactory accuracy. The missing component to this array of research has been a study of the recorded sunshine duration at the first-order stations in Iran.

Based on Blani Kridel's equation, the theoretical maximum possible sunshine hours are rather high over the country with a maximum of roughly 15 h per day on 22 June and a remarkably high minimum of 9.5 h per day on 22 December ( (Black, 1954). These values are needed to draw attention to the different aspects of solar energy's resource potential to estimate probable returns on investments including temporal and spatial variability of sunshine hours.

The purpose of this paper is to illustrate the temporal and spatial variability and trends of sunshine duration in different geographical regions in Iran. Section '2. Data' is dedicated to the datasets used: their period and selected sites. Two simple used methods including box plot and linear trend analysis are described briefly in Section '3. Methods'. The fourth section discusses the classification and trends of the original time series and how it was segmented into four unique sub-regions. Finally, the conclusion and summary of this paper is provided in Section '5. Conclusion'.

2. Data

In Iran, routine measurement of the sunshine duration using the Campbell Stokes Ball began in 1951 at a few synoptic stations belonging to the meteorological network of IRIMO (Islamic Republic of Iran Meteorological Organization). Since then, the number of sites has increased to over 350. As may be expected, there are some issues in the historical records such as missing data, coarse discontinuity and different periods, probable relocation or environmental changes of the stations, and non-systematic errors and instrument changes. The related problems were probed deeply in the project ‘Statistical Detection of Climate Change in Iran for period 1960–2005’ by ASMERC.

In the case of sunshine data, although the longest period of records is 1951–2007, it is limited to just a few stations. One of the encompassing problems relates to missing data during 1976–1982 for some Iranian synoptic stations. As a result, to gain better coverage and also for equity of period for better judgment of spatial changes, 37 stations with over 95% completeness during the period 1981–2007 were used in this study. The geographical characteristics of the 37 synoptic stations are shown in Table 1.

Table 1. Characteristics of selected synoptic stations for examination of sunshine hours trend over Iran
No.StationGeographical co-ordinatesNo.StationGeographical co-ordinates
Latitude (N)Longitude (E)Elevation (m)Latitude (N)Longitude (E)Elevation (m)
1Anzali49 °28 ′37 °28 ′− 2620Khoy38 °33 ′44 °58 ′1103
2Ahvaz31 °20 ′48 °40 ′1821Kish26 °30 ′53 °59 ′30
3Arak34 °06 ′49 °46 ′170822Mashhad36 °16 ′59 °38 ′990
4Babolsar36 °43 ′52 °39 ′− 2123Noshahr36 °39 ′51 °30 ′− 21
5Bam29 °09 ′58 °21 ′106724Oroomieh37 °32 ′45 °5 ′1312
6Bandar Abbas27 °13 ′56 °22 ′1025Ramsar36 °54 ′50 °40 ′− 20
7Birjand32 °52 ′59 °12 ′149126Rasht37 °12 ′49 °39 ′37
8Bojnourd37 °28 ′57 °19 ′109127Sabzevar36 °1257 °43 ′978
9Bushehr28 °59 ′50 °50 ′2028Sanandaj35 °2047 °00 ′1373
10Esfehan32 °40 ′51 °52 ′160129Shahrekord32 °20 ′50 °51 ′2061
11Fasa28 °58 ′53 °41 ′128830Shahroud36 °05 ′54 °57 ′1345
12Ghazvin36 °15 ′50 °00 ′127831Shiraz29 °33 ′52 °36 ′1488
13Gorgan36 °51 ′54 °16 ′1332Tabriz38 °5 ′46 °17 ′1361
14Hamedan-Nojeh35 °12 ′48 °43 ′168033Tehran-Mehrabad35 °41 ′51 °19 ′1191
15Hamedan- Foroudgah34 °52 ′48 °32 ′174234Yazd31 °5454 °24 ′1230
16Iranshahr27 °12 ′60 °42 ′59235Zabol31 °13 ′61 °29 ′489
17Kerman30 °15 ′56 °58 ′175436Zahedan29 °28 ′60 °53 ′1369
18Kermanshah34 °19 ′47 °7 ′132237Zanjan36 °41 ′48 °41 ′1663
19Khoramabad33 °29 ′48 °22 ′1125     

3. Methods

A box plot of sunshine hours was generated to show the maximum, minimum, 25th and 75th percentiles as well as the mean sunshine hours. By combining a box plot from all stations the following questions may be answered: (1) is the distribution of annual sunshine hours nearly symmetric, (2) how does the mean and variation in annual sunshine hours differ between stations, (3) are there any outliers for sunshine hours observation, and, (4) what are the implications of changing sunshine hours in Iran?

Cluster analysis was used to identify relatively homogeneous groups of stations in terms of variability of sunshine during 1981–2007, using an algorithm k-mean cluster analysis to identify relatively homogeneous groups of stations based on their seasonal sunshine hours. Using the box plot analysis and the range of annual and seasonal sunshine duration values, and the result of cluster analysis, four distinct regions were identified over the country. By averaging the time series of each region stations and standardizing them, four regional representative time series were also developed. These proxy series were used to study the variations of sunshine duration in each region.

Further, trends rate (â) and estimated correlation co-efficient (r̂) for a linear trends (Y = aX + b) and also were calculated by using least-square regression methods using n pairs of observations with given co-ordinates (Xi, Yi). There are different approaches to make an inference about the correlation co-efficient affected or not affected by serial correlation (von Storch and Zwiers, 1999). Following no significant serial correlation, and according to the constant length period where n = 27, the accepted level of significance for a linear trend at the 5% level for n = 27, when r̂ ≥ 0.367.

4. Results

4.1. Regionalization

The cluster analysis show, existence of four clusters that are significant according to statistical tests. The number of iterations for achieving convergence was 3 and the absolute co-ordinate change for any cluster was 0.0 at the end of the iterations. The minimum distance between initial centres is 40.8. The F tests that were used for descriptive purposes to maximize the difference among stations in different clusters are 156.8, 110.0, 108.0 and 116.3 for spring, summer, autumn and winter, respectively, with 33 degrees of freedom and consequently acceptance of them at the 0.0 significance level. Using the result of cluster analysis on time series of sunshine hours, comparing the time series of sunshine hours derived from the selected stations, accompanied with the box plots (Figures 1 and 2), four distinct sub-regions were determined for Iran (Table 2, Figure 3), as follows.

  • Region I: a narrow strip of land located on the southern side of the Caspian sea (six stations). Table 2 shows that the geographical and seasonal variability of the average monthly mean sunshine hours are different within the Caspian sea region versus other regions of the country. The maxima, minima, mean, and 25th and 75th percentile measures of the stations located in this side of Caspian Sea region have the lowest values. The Ramsar station recorded the shortest hours of 1575.1 h per year (∼4 h per day). The high mountains to the west, south and east of this small coastal region act as a dam, preventing the inland penetration of warm and moist air from the Caspian Sea. There isn't any obstruction at the northern side of the area and thus it is open to atmospheric currents. The movement of cold high pressure systems over the warm sea, as well as north–south currents, helps to produce a nearly daily sea-breeze circulation. Consequently, most of the time, the Caspian Sea southern coastal region is cloudy and/or rainy. Because the flow of cold air over the warmer sea is generally initiated from the west, the cloudiness and rainfall intensity of this part is greater than its eastern counterpart. As the cold air moves toward the east, the cloudiness and rainfall intensity eventually decrease. For instance, according to Table 2, Ramsar in the west has the lowest annual mean sunshine hours duration and Gorgan, in the east, has the maximum value (2150.9 to 5.8 h per day).

  • Region II: vast mountainous area including west and northwest, east and northeast (18 stations). The duration of sunshine hours in this area is roughly equal in the west, northwest and northeast parts of the country. These areas are exposed to a variety of differing, and even intense, weather systems (rain, snow, sunshine), especially in autumn and winter. Among the stations that are located here, the longest hours during summer were observed at Mashhad (1036.1 h).

  • Region III: central area in margins of Kavir (nine stations). The lower elevation of the central region is typically dry and sunny. As the weather systems have usually ‘filled’ or dissipated before reaching this area, there aren't many synoptic-scale factors for cloud formation. The Bam station located in the central area registered the longest average annual sunshine hours of about 3387 h per year (∼9 h per day). The longest hours during spring is found at Fasa with an average of 949.0 h, and in winter the longest hours were found at Iranshahr, with a mean of 711.0 h.

  • Region IV: southern region including area across Persian Gulf and Oman sea (four stations). The seasonal and annual sunshine hours for the south regions are less than for Region III for the whole of the year, and the sunshine hour duration for the autumn and winter seasons are remarkably higher with respect to the northern half of the country. For the station at Iranshahr, in the average number of sunshine hours in autumn and winter 711.0 and 810.6 h, the highest in the region.

Figure 1.

Box plot for annual sunshine hours for 37 stations following Wilks (2006). The ‘outside’ values including ‘○: inner fence’ and ‘*: out fence’ are plotted in accompanied by the number of records in the seasonal sunshine hours at the right side of symbols separately. This figure is available in colour online at

Figure 2.

Box plot for seasonal sunshine hours ((a) spring, (b) summer, (c) autumn, and (d) winter) for 37 stations following Wilks (2006). The ‘outside’ values including ‘○: inner fence’ and ‘*: out fence’ are plotted in accompanied by the number of records in the seasonal sunshine hours at the right side of symbols separately. This figure is available in colour online at

Figure 3.

Four segmented sub-regions of the country in terms of sunshine hours. This figure is available in colour online at

Table 2. Seasonal mean of sunshine hours segmented into four regions for the period 1981–2007
  1. a

    Lowest value in the season/annual.

  2. b

    Highest value in the season/annual.

 Hamedan- Foroudgah862.71001.3562.4507.32933.7
 Bandar Abbas870.6800.4745.0651.83060.9

In general, the sunshine hours in the central and southern regions are higher than in the north regions, save for the Caspian region. The spatial variability over the country is magnified during autumn and winter. In fact all areas in the west, northwest and northeast have a similar pattern, and the only difference is that the sunshine duration in the autumn and winter have relatively larger values for central Iran.

Figure 4 shows the similar behaviour of monthly sunshine hours in the four defined regions except for Region IV during June to October.

Figure 4.

Mean monthly sunshine hours time series during 1980–2005. This figure is available in colour online at

4.2. Trend analysis

An examination of the change rates in sunshine hours for each of the 37 stations was conducted for the 1981–2005 period. This analysis was performed using seasonal and annual mean values of the observed sunshine hours data. Schematically, Figures 5 and 6 demonstrate spatial decadal trends in seasonal and annual sunshine hours duration in accompanied by to be or not significant at 0.05 level for selected stations (hours per decade) for the 1981–2007 period over Iran. The result of trend analysis show that the mean annual sunshine hours have increased at most of the stations used in this study. Almost all of the annual time series plots, that is 31 of the 37 individual annual time series examined, showed statistically significant (p ⩽ 0.05) trends. The trends were highly significant (α = 0.05) for 18 cases in spring, 26 cases in summer, 20 cases in autumn, 18 cases in winter. The greatest rates of change rates were found in the mountain area for Sanandaj (spring, summer), Khoy (autumn) and Hamedan-Foroudgah (winter) and annual annually (Sanandaj), and lowest amounts relate to Zabol (spring and summer), Kish (autumn), Bushehr (winter) and Bam (year).

Figure 5.

Decadal trends in seasonal sunshine hours (h per decade) for the 1980–2005 period over Iran. (a) spring, (b) summer, (c) autumn, (d) winter. This figure is available in colour online at

Figure 6.

Decadal trends in annual sunshine hours (hours per decade) for the 1980–2005 period over Iran. This figure is available in colour online at

Figure 7(a) demonstrates the average annual variations of sunshine duration for each region, according to the four representative series. These representatives were standardized for clarification of the signals in time series. As the selected period is 27 years, they were standardized with respect to a 10 year period (1981–1990). Figure 7(b) shows the mean standardized values of sunshine duration. The annual patterns for the four regions show remarkably similar features. For example, there was a small increase in the annual sunshine duration in the early 1990s across all regions. Simultaneously, the most remarkable feature (Figure 7(b)) is a large reduction in sunshine duration over all regions, including the least sunny Caspian Sea, with relatively lower sunshine duration after the years of 1990 and 2003. Relatively low values of standardized sunshine hours in three individual short periods, 1980–1985, 1990–1995 and 2000–2005, is experienced for the whole series.

Figure 7.

(a) Annual time series representative of each four regions. (b) Standardized values of time series representative of each four region accompanied with number of sunspot index in secondary vertical axis. This figure is available in colour online at

According to Table 3, the linear trend for the time series of regions are significant during the period at 0.05 confidence interval but equally apparent is the nonlinear behaviour. The mean Iran sunshine duration shows clear, nonlinear, long-term fluctuations (Figure 7(b)) and an existence of a significant periodic component in the distribution of the annual totals of sunshine duration. Maximum curve peaks have been observed in 1989, 2000 and minima in 1992, 2003. Four time series graphs, corresponding to each of the four regions, show a non-significant (P ≥ 0.05) correlation between the annual sunshine hours and the sunspot index number (SSN) (Table 3). Figure 7(b) demonstrates that some alignment in the data exists between peak value for time series of sunshine hours and sunspot index.

Table 3. Trends characteristics of each four region and statistical correlations between annual sunshine duration and sunspot index number
RegionsLinear relation between sunspot and sunshine hoursTrends characteristics
I− 0.0860.02613.6250.666
III− 0.4360.16311.9550.716
IV− 0.6210.21911.7110.539

5. Conclusion

It is essential to understand the potential of the solar energy resource in different regions of Iran due to prognosticated changes of major climate variables, including cloud cover, duration of sunshine hours and solar flux at the Earth's surface. Due to the geographical location of Iran, the long-term mean sunshine duration is relatively high. As such, a rather high potential for solar energy production exists. Through the inference of the impact of climate change on sunshine duration, such as some other climatic factors, the spatial and temporal variability of this parameter was examined during the period 1981–2007, at 37 first-order synoptic sites of the Iran meteorological station network.

Firstly, 37 unique and individual time series of sunshine duration were compared throughout the period using the box plot method. From these plots, and in terms of those annual and seasonal values, the distribution of sunshine hours could be sub-divided into four regions: the southern side of the Caspian sea, the mountainous area at the west and east, the central area in the margins of Kavir and the area across Persian Gulf and Oman sea. Four regional representative time series were developed through averaging of the time series of sunshine hours in each subregion and standardizing them. Despite the fact of a large reduction over all regions, especially after 1990 and 2003, there was still a noticeable increase in sunshine duration during the period 1981–2007 on the annual and seasonal scales. These results are consistent with positive trends at some sites in the northwest quarter of the United States of America and Tibet and in contrast of negative trend for the other probed regions over the world (Stanhill and Cohen, 2005; Chen et al., 2006; Yang et al., 2009). Furthermore, the mean Iran sunshine duration indicates a significant periodic component for the country. Curve peaks in the two cases have observed maxima (for 1989, 2000) and minima (1992, 2003).

Given the increasing trends observed at many stations in Iran over the last 25 years, and recognizing the natural high sunshine duration experienced across the region, it is believed that a strong case can be made for the introduction of solar energy across the country. Peak sunshine hours regularly exceed 9 h per day in high latitude locations. It also understood that while solar energy is effective virtually anywhere, the desired minimum amount of sunshine needed to generate effective energy from solar panels is roughly 4–6 h per day. As alternative energy solutions continue to rise in an effort to offset carbon emissions from traditional energy generators, the country of Iran appears poised to participate in this effort through solar energy production.


The authors wish to thank Mahnaz Khazaie for her support and production of the maps included in this manuscript. The authors also thank several anonymous reviewers for their suggestions toward improving this manuscript.