Bilingual
En/Fa
Geographical Research is Published in both Persian and English Full-text.
Volume 37, Issue 4 (2022)
GeoRes 2022, 37(4): 529-540 |
Back to browse issues page
Article Type:
Subject:
History
Received: 2022/10/11 | Accepted: 2022/11/30 | Published: 2022/12/16
Received: 2022/10/11 | Accepted: 2022/11/30 | Published: 2022/12/16
How to cite this article
Derikvand S, Nasiri B, Ghaemi H, Karampoor M, Moradi M. Statistical Analysis of High- and Low-Precipitation Areas of Iran. GeoRes 2022; 37 (4) :529-540
URL: http://georesearch.ir/article-1-1395-en.html
URL: http://georesearch.ir/article-1-1395-en.html
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Rights and permissions
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
1- Department of Geography, Faculty of Literature and Humanities, Lorestan University, Khorramabad, Iran
2- Research Institute of Meteorology and Atmospheric Science, Tehran, Iran
2- Research Institute of Meteorology and Atmospheric Science, Tehran, Iran
Full-Text (HTML) (126 Views)
Introduction
Precipitation is a phenomenon arising from complex atmospheric interactions and holds critical importance due to its essential role in environmental and human systems. Iran is among the regions that experience irregular and unpredictable precipitation patterns [Mohammadi, 2010]. Although precipitation is fundamentally important in every climate, it plays an especially vital role in the daily lives of people living in arid and semi-arid regions such as Iran [Masoudian & Kavian, 2008]. Due to its specific climatic and geographical conditions, Iran is inherently a low-rainfall region. The country’s average annual precipitation is only slightly above 250 mm; therefore, Iran is generally classified among the relatively dry regions of the world. Moreover, this already limited rainfall is unevenly distributed across the country [Sedaghat, 2001].
Heavy precipitation events cause severe human and financial losses each year in Iran and around the world and are considered a major natural hazard, often leading to disasters such as storms and floods [Davarpanah, 2004]. Between 1990 and 2007, Iran recorded 24 occurrences of hazardous heavy rainfall, resulting in 13 deaths, 9 injuries, and approximately 41 billion rials in financial damages [Farajzadeh, 2012]. Conversely, below-average rainfall leads to drought, which causes substantial agricultural losses, declining groundwater levels, reduced reservoir storage, and consequent disruptions in services dependent on dam operations. Drought severity varies regionally and from year to year. Over an 18-year study period, Iran experienced 35 hazardous drought events, directly affecting around 110,000 people; the financial losses associated with drought were four times greater than the combined losses from all other studied hazards [Farajzadeh, 2012].
These conditions highlight the necessity of identifying widespread heavy precipitation events as well as extensive low-rainfall periods, assessing their monthly probability of occurrence, and determining the regions most affected by these extremes. To examine the potential influence of atmospheric systems or targeted phenomena on national precipitation patterns, it is first essential to statistically identify anomalous precipitation dates. Thus, extracting the dates of maximum and minimum precipitation, the most widespread heavy rainfall events, the largest low-rainfall areas, periods of consecutive dry or wet years, and long-term precipitation variability is of critical importance. Knowing when these extreme conditions occur facilitates more precise investigation of their underlying causes.
A substantial body of research has addressed various aspects of precipitation through statistical analyses, each aligned with its specific research objectives. Using clustering techniques, precipitation extremes in Iran have been classified into eight homogeneous regimes, and wet and dry half-years have been identified for each regime [Mohammadyarian et al., 2019]. An examination of precipitation variability from 1961 to 2003 reveales that approximately 51.4% of Iran’s land area experienced significant precipitation changes, particularly in mountainous and western regions. Areas with higher rainfall generally exhibited greater variability, with annual changes ranging from +29.6 mm in Kouhrang to –15.7 mm in Sarab [Asakare, 2007].
A 50-year analysis of effective precipitation in Iran showed that the median effective rainfall at selected stations is negative, indicating a general decreasing trend. Similarly, the median total precipitation for the 40-year period 1961–2000 was near zero, whereas for 1971–2010 it was significantly negative at the 10% level. Although the median for the full 50-year period (1961–2010) was also negative, it was not statistically significant [Torabipoodeh et al., 2018]. Statistical distribution assessments for extreme rainfall risk mapping in western Iran identified the log-logistic, Pearson, and gamma distributions as the most suitable models [Salimi Mostala et al., 2019]. Over the past six decades, rainfall patterns in Iran have changed noticeably: Rainy days have decreased, while maximum rainfall events and days with heavy precipitation have increased significantly. However, total annual rainfall has fluctuated without exhibiting a consistent upward or downward trend across the country [Doostan, 2020].
Rainfall variability in the Kashafrud basin indicates no significant trend in annual precipitation, although winter rainfall showed a decreasing trend and autumn rainfall in the southern half exhibited an increasing trend. Sharp decreases in winter precipitation, particularly in December and January, pose serious challenges for water resource management during the dry season [Khosravi et al., 2022]. In studies of extreme rainfall in western Iran using threshold approaches (mean precipitation above the 75th percentile plus multiples of its standard deviation), three categories of extreme events were identified, and synoptic patterns corresponding to each spatial extent were extracted [Jahanbakhsh Asl et al., 2022].
Fractal analysis is among the modern approaches used to study precipitation characteristics. In the Karkheh and Dez basins, Pearson correlation-based fractal analysis revealed no significant relationship between precipitation regimes and topography in the Karkheh basin across spatial scales; however, in the Dez basin, a significant relationship existed at the small-scale regime, but not at the large-scale regime, likely due to heterogeneous and contrasting topographic conditions in the Karkheh basin [Hasanvand et al., 2022].
Precipitation trends vary globally across temporal and spatial scales and are strongly influenced by other climatic parameters. Hourly precipitation analysis in Côte d’Ivoire over a 10-year period showed a negative trend in extreme hourly rainfall. Studies linking floods with atmospheric circulation patterns indicate that flood-producing precipitation in different geographic regions is associated with distinct synoptic systems, many of which have now been identified [Soro et al., 2016].
Investigation of extreme rainfall in China using satellite imagery and multiple Doppler weather radars revealed that catastrophic heavy rainfall resulted from a mid-level convective system around 14 km altitude. Local convergence driven by regional topography and the intrusion of a cold tongue at the mountain base intensified the rainfall system [Li et al., 2020]. Additional analyses using principal component methods identified six major weather patterns and three key centers of heavy rainfall. These studies also demonstrated that the spatial occurrence of large-scale heavy rainfall events was closely related to urban density, indicating that urbanization significantly influences the magnitude, intensity, and spatial concentration of such events [Li et al., 2021].
The objective of the present study was to extract the key dates associated with precipitation variability, not only the extreme wet and dry events but also the spatial extent involved, and to identify temporal changes in precipitation throughout the study period on a month-by-month basis.
Methodology
This study employed daily precipitation data from 117 synoptic stations across Iran with a shared statistical baseline for the period 1986–2020. These data were obtained from the Data Provision Department of the Iran Meteorological Organization. Because most precipitation in Iran occurs during autumn, winter, and spring, the analysis focused on the period from October to June of each year.
Homogeneity testing was performed using XLSTAT (Standard Normal Homogeneity Test, SNHT), an add-in for Microsoft Excel. Stations with the longest common statistical period and sufficient data continuity were selected. Prior to 1986, the number of synoptic stations was too limited to allow reliable spatial interpolation; and although the number of stations increased after this date, the available record still does not constitute a long-term climatological baseline. A larger number of stations with balanced spatial distribution leads to more accurate interpolation and more reliable assessment of spatial and temporal precipitation variations. Ultimately, among the available stations, those with more complete and reliable meteorological records were selected for analysis. No data reconstruction was applied in this study. Despite the increase in the number of synoptic stations in recent years, their spatial distribution remains heterogeneous and not uniformly distributed across the country. Monthly mean precipitation for each station was calculated using the arithmetic mean. To determine whether each month at each station was wetter or drier than average, the precipitation of each month was compared with the long-term monthly mean. The resulting values were categorized as positive, zero, or negative. A positive value indicates precipitation above the long-term mean; zero indicates equality with the mean; and a negative value indicates lower-than-average precipitation.
The deviations for all stations were calculated using MATLAB 2013, and the interpolated spatial patterns were generated using SURFER 24.3 to identify areas experiencing above-average or below-average precipitation, as well as the magnitude of these deviations.
A dual-line chart was constructed to illustrate the number of wet stations (precipitation above the mean) and dry stations (precipitation below the mean) in each month. The inverse relationship between these two categories, where an increase in one corresponds to a decrease in the other, helps identify the overall spatial patterns of wetter or drier conditions during the study period. The distance between the two lines and their points of intersection provide insight into the dominance and temporal shifts of these precipitation regimes.
Findings
Mean Precipitation and Variability of Precipitation Zones
The mean precipitation from October to June during the period 1986–2020 and the variations in precipitation zones were examined.
October: The highest precipitation levels in this month occurred in the northern regions, reaching up to 260 mm, followed by the northwest and the Kouhrang area of Isfahan with more than 120 mm. The southernmost boundary of precipitation above 20 mm extended to the southern parts of Lorestan Province. However, a vast portion of the country received less than 20 mm of rainfall. Two seven-year dry periods were observed during 1994–1998 and 2001–2008. Only in six years did the number of wet stations exceed that of dry ones. In recent years, the number of wet Octobers has increased slightly.
November: The area receiving more than 20 mm of precipitation expanded compared to October and reached the southwestern regions. The northern parts of the country still recorded the highest precipitation, around 260 mm. Kouhrang, with more than 120 mm of rainfall, had the highest precipitation after the Caspian Sea coasts. Western, northwestern, and northeastern regions also recorded more than 80 mm. At the beginning of the study period, an eight-year dry spell was evident. In eleven years, the number of wet stations exceeded dry ones, mostly in the later years.
December: In this month, the western, southwestern, and northern regions received comparatively more rainfall. Although the spatial distribution of precipitation showed little change compared to November, the amount recorded at stations generally increased. Southern regions, parts of the central plateau, and the northeast received more than 50 mm, whereas the east, southeast, and central parts had lower averages. A wet period was observed from 2000 to 2004. In ten years, the number of wet stations exceeded the dry ones. Unlike October and November, dry zones increased in recent years.
January: Rainfall in January expanded further eastward compared to December. The western to northwestern regions received higher rainfall, and precipitation extended farther south. The lowest precipitation occurred in central and eastern regions, with the highest levels in the southwest, west, and north. Several short and long dry periods appeared in recent years, and in twelve years, dry zones exceeded wet ones.
February: The maximum precipitation increased from 180 mm in January to 200 mm, mainly in western regions and the western Caspian coast. Rainfall decreased in the east and increased in central areas. Other stations experienced little change. At the beginning of the study period, wet and dry areas were similar in extent, but in recent years dry zones surpassed wet ones in both number and spatial coverage. Only in six years did wet stations outnumber dry ones.
March: Peak precipitation shifted from the Caspian coastal regions to the southwestern heights and Kouhrang. Areas receiving less than 30 mm expanded. Most stations recorded lower precipitation except for limited regions such as the central Zagros elevations, where precipitation increased. Patterns of dry and wet periods were clearly visible, and recent years did not differ significantly from earlier periods. In seven years, above-average precipitation covered a larger area of the country.
April: Rainfall patterns changed noticeably compared to March. Peak precipitation remained in Kouhrang but decreased by nearly 100 mm. Central and southeastern regions recorded almost no precipitation, whereas northwest, west, and parts of the southwest received up to 70 mm. Large central areas recorded up to 20 mm. An eight-year dry spell occurred from 2000 to 2008. In twelve years, above-average rainfall covered extensive areas, with broad and heavy precipitation reported in recent years.
May: Peak precipitation occurred in the northwest, reaching up to 55 mm. The lowest precipitation (up to 5 mm) was observed in central, southern, and southeastern regions. Only in four years did wet zones exceed dry ones, and widespread dryness dominated recent years.
June: Rainfall declined sharply compared to May. The highest precipitation occurred in the northern regions but did not exceed 46 mm, although these regions were the wettest relative to the rest of the country. Southern Sistan and Baluchestan saw a slight increase and was the only region with higher rainfall than in the previous month. Most of the country received less than 4 mm. Only in three years did wet stations slightly outnumber dry ones. In recent years, a pronounced dry period has developed, with dry zones far more extensive than wet ones.
Wettest and Driest Months Based on Spatial Extent
The wettest October, driest November, and driest January during the 35-year period all occurred in 1987, indicating that this year experienced the most expansive wet and dry zones. January and February of 1996 had the largest areas with above-average precipitation. In March and May 2008, the widest wet zones were recorded.
Absolute dryness with below-average rainfall across all 117 stations occurred in November and January of 1987. The highest number of stations with above-average rainfall, 102 stations, was recorded in March 1996. December and January had the broadest spread of above-average precipitation; in 15 years, more than half of the stations experienced above-average rainfall. In June, only three years had more than half of the stations recording above-average precipitation, while in the remaining years, most stations had below-average rainfall.
Discussion
In addition to the statistical analysis of precipitation, the objective of this study was to determine the spatial extent of areas experiencing below- or above-average precipitation from October to June. The spatial coverage of above-average precipitation also reflects the extent of the precipitation-inducing systems. Identifying the years with the widest precipitation coverage for each month provides a basis for examining the synoptic systems responsible for such events.
Assessing monthly precipitation anomalies relative to the same month in other years is important because, when analyzing the temporal pattern of precipitation across consecutive months, climatic factors such as solar angle remain constant. However, large-scale systems, whose strength depends on the solar angle, do not behave uniformly across the months of a single year. Thus, comparing each month to its own long-term average while holding constant the factors that do not change allows us to identify additional causes that make each month unique relative to its climatological norm. In this way, each month’s precipitation is compared only with its own type, and not with months that do not share similar atmospheric conditions.
Using the dual-line diagram applied in this study, the spatial extent of heavy precipitation for each month can be clearly observed and quantified, along with the magnitude of precipitation deficits. Moreover, changes in the spatial coverage of heavy precipitation from year to year for each month are identifiable. The strength and expansion of above-average precipitation vary noticeably from one month to another. For example, in autumn the spatial extent of above-average precipitation is smaller than in winter. Additionally, different regions of the country enter the category of high-precipitation zones in different months; for instance, the eastern and southeastern regions tend to be included in the wet zone at times when most other regions fall outside it. Overall, analysis of the dual-line diagrams provides extensive information about the precipitation characteristics of each month. Identifying the wettest and driest years, along with the spatial extent of these conditions for each month during the study period, is one of the distinguishing features of this research.
Previous studies have examined precipitation trends and variability across different regions of the country using the Mann–Kendall test. This test has been applied to evaluate the significance of precipitation trends, spatial patterns of rainfall amounts, the highest and lowest coefficients of variation, and the frequency of dry and wet periods in the study regions [Mirian et al., 2022; Mazidi et al., 2022; Amani et al., 2020]. Markov chains have also been widely used as a stochastic model for describing sequences of probabilistic events in which the probability of each event depends solely on the previous state. Numerous studies have used this approach to analyze precipitation trends and ultimately forecast rainfall. Analyses have included the persistence and frequency of wet and dry days, the probability of rainfall occurrence, rainfall amounts, and return periods. Time-series trends of wet and dry periods, their intensities, and the probability of transitions between wet and dry states have also been investigated using this method [Ghaznavi et al., 2021; Iranpour et al., 2019].
Drought monitoring can be carried out from climatological, hydrological, agricultural, economic, and social perspectives. In climatological drought assessments, the Standardized Precipitation Index (SPI) is widely used. The SPI at 12-, 24-, and 48-month timescales has been calculated for selected stations, and interpolated maps have been produced. Moran’s spatial autocorrelation statistic and hot-spot analysis have been used to identify drought patterns and map their distribution. Findings for Fars Province indicate that nearly all parts of the province experience drought with varying severities. Results also show that as the temporal scale increases, SPI values rise, while the absolute magnitude of drought decreases [Ghanbari, 2020]. The SPI is also used to identify wet and dry conditions in different regions and is extensively applied for drought monitoring and forecasting. In this method, wetness and dryness levels are classified into grades with defined thresholds [Bazrafshan et al., 2011; Shokohi et al., 2018; Mohtashami et al., 2015].
Changes in the waiting time between rainfall events have also been studied. Using geostatistical kriging, daily precipitation values were interpolated on 6×6 km grid cells. A non-parametric Mann–Kendall test at the 90% confidence level was applied to the time series of waiting times between precipitation events, and rate of change was estimated using linear least squares. Findings indicated that the waiting time between rainfall events has increased during the cold months of the year, whereas the intervals between consecutive rainfall events have shortened during the warm months [Darand et al., 2019].
Monthly rainfall behavior at spatial and temporal scales has also been examined through the relationship between precipitation and topographic parameters. Although a statistically significant relationship exists between monthly rainfall variability and spatial–topographic factors, correlations have not been particularly strong. Using cluster analysis, five regions were identified. These regions generally extend northward from latitude 32°N, while the area south of 32°N tends to follow longitudinal patterns [Asakare et al., 2021]. Heavy precipitation has also been analyzed using a threshold of the 95th percentile and non-parametric tests such as Mann–Kendall, Sen’s estimator, and simple linear regression. Results indicated a significant increasing trend only at the Astara station, while other stations showed non-significant upward trends [Zeynali et al., 2020].
In the present study, only one threshold, the long-term mean, was used to classify each station into wet (above-average precipitation) or dry (below-average precipitation). This binary classification facilitates the analysis of related parameters, such as the duration of wet and dry periods, the spatial differences between wet and dry areas, and their temporal variability. Using multiple thresholds would complicate the classification process and make it more difficult to analyze the relationships among categories. In the Mann–Kendall test and Markov chain analysis, if precipitation deviates substantially from normal conditions, trend significance may no longer be detectable, which renders forecasting with these methods ineffective. Given the influence of climate change and global warming, an increase in non-normal precipitation events is expected. Another consequence of climate change is the spatial–temporal displacement of precipitation patterns, a highly parameter and chaotic atmospheric parameter. While previous studies have assessed precipitation intervals in warm and cold seasons using linear least squares, the present study identifies intervals of wet and dry conditions for each month across consecutive years. This study does not aim to analyze precipitation trends or provide forecasts; rather, it focuses on extracting key spatiotemporal variations in precipitation to facilitate the investigation of their underlying causes.
One of the limitations of this study is its use of a single threshold, which does not allow for grading the severity of wetness or dryness in the study areas. Future research may incorporate multiple precipitation thresholds appropriate to regional climates if the research objectives require such refinement. Finally, given the identification of the driest and wettest months, as well as the years with the widest spatial extent of dryness and heavy precipitation, it is recommended that future studies complement these results with synoptic and dynamical analyses of the responsible atmospheric systems.
Conclusion
Precipitation in Iran generally occurs over relatively limited spatial extents. Although below-average precipitation is observed across all stations at times, there has never been a case in which all stations across the country experience above-average precipitation simultaneously.
Acknowledgments: We sincerely thank the Iran Meteorological Organization for providing the data used in this study.
Ethical Permission: No ethical issues were applicable.
Conflict of Interest: The authors declare no conflicts of interest.
Author Contributions: Derikvand S (First Author), Main Researcher (50%); Nasiri B (Second Author), Discussion Writer (15%); Ghaemi H (Third Author), Methodologist (15%); Karampoor M (Fourth Author), Data Analyst (10%); Moradi M (Fifth Author) – Introduction Writer (10%)
Funding: This study is derived from the first author’s doctoral dissertation titled “Investigation of the Impact of the Overlying Atmosphere on the Lower Atmosphere in Intensifying or Weakening Iran’s Precipitation during the Cold Season”, conducted at the University of Lorestan under the supervision of the second and third authors and with consultation from the fourth and fifth authors.
Precipitation is a phenomenon arising from complex atmospheric interactions and holds critical importance due to its essential role in environmental and human systems. Iran is among the regions that experience irregular and unpredictable precipitation patterns [Mohammadi, 2010]. Although precipitation is fundamentally important in every climate, it plays an especially vital role in the daily lives of people living in arid and semi-arid regions such as Iran [Masoudian & Kavian, 2008]. Due to its specific climatic and geographical conditions, Iran is inherently a low-rainfall region. The country’s average annual precipitation is only slightly above 250 mm; therefore, Iran is generally classified among the relatively dry regions of the world. Moreover, this already limited rainfall is unevenly distributed across the country [Sedaghat, 2001].
Heavy precipitation events cause severe human and financial losses each year in Iran and around the world and are considered a major natural hazard, often leading to disasters such as storms and floods [Davarpanah, 2004]. Between 1990 and 2007, Iran recorded 24 occurrences of hazardous heavy rainfall, resulting in 13 deaths, 9 injuries, and approximately 41 billion rials in financial damages [Farajzadeh, 2012]. Conversely, below-average rainfall leads to drought, which causes substantial agricultural losses, declining groundwater levels, reduced reservoir storage, and consequent disruptions in services dependent on dam operations. Drought severity varies regionally and from year to year. Over an 18-year study period, Iran experienced 35 hazardous drought events, directly affecting around 110,000 people; the financial losses associated with drought were four times greater than the combined losses from all other studied hazards [Farajzadeh, 2012].
These conditions highlight the necessity of identifying widespread heavy precipitation events as well as extensive low-rainfall periods, assessing their monthly probability of occurrence, and determining the regions most affected by these extremes. To examine the potential influence of atmospheric systems or targeted phenomena on national precipitation patterns, it is first essential to statistically identify anomalous precipitation dates. Thus, extracting the dates of maximum and minimum precipitation, the most widespread heavy rainfall events, the largest low-rainfall areas, periods of consecutive dry or wet years, and long-term precipitation variability is of critical importance. Knowing when these extreme conditions occur facilitates more precise investigation of their underlying causes.
A substantial body of research has addressed various aspects of precipitation through statistical analyses, each aligned with its specific research objectives. Using clustering techniques, precipitation extremes in Iran have been classified into eight homogeneous regimes, and wet and dry half-years have been identified for each regime [Mohammadyarian et al., 2019]. An examination of precipitation variability from 1961 to 2003 reveales that approximately 51.4% of Iran’s land area experienced significant precipitation changes, particularly in mountainous and western regions. Areas with higher rainfall generally exhibited greater variability, with annual changes ranging from +29.6 mm in Kouhrang to –15.7 mm in Sarab [Asakare, 2007].
A 50-year analysis of effective precipitation in Iran showed that the median effective rainfall at selected stations is negative, indicating a general decreasing trend. Similarly, the median total precipitation for the 40-year period 1961–2000 was near zero, whereas for 1971–2010 it was significantly negative at the 10% level. Although the median for the full 50-year period (1961–2010) was also negative, it was not statistically significant [Torabipoodeh et al., 2018]. Statistical distribution assessments for extreme rainfall risk mapping in western Iran identified the log-logistic, Pearson, and gamma distributions as the most suitable models [Salimi Mostala et al., 2019]. Over the past six decades, rainfall patterns in Iran have changed noticeably: Rainy days have decreased, while maximum rainfall events and days with heavy precipitation have increased significantly. However, total annual rainfall has fluctuated without exhibiting a consistent upward or downward trend across the country [Doostan, 2020].
Rainfall variability in the Kashafrud basin indicates no significant trend in annual precipitation, although winter rainfall showed a decreasing trend and autumn rainfall in the southern half exhibited an increasing trend. Sharp decreases in winter precipitation, particularly in December and January, pose serious challenges for water resource management during the dry season [Khosravi et al., 2022]. In studies of extreme rainfall in western Iran using threshold approaches (mean precipitation above the 75th percentile plus multiples of its standard deviation), three categories of extreme events were identified, and synoptic patterns corresponding to each spatial extent were extracted [Jahanbakhsh Asl et al., 2022].
Fractal analysis is among the modern approaches used to study precipitation characteristics. In the Karkheh and Dez basins, Pearson correlation-based fractal analysis revealed no significant relationship between precipitation regimes and topography in the Karkheh basin across spatial scales; however, in the Dez basin, a significant relationship existed at the small-scale regime, but not at the large-scale regime, likely due to heterogeneous and contrasting topographic conditions in the Karkheh basin [Hasanvand et al., 2022].
Precipitation trends vary globally across temporal and spatial scales and are strongly influenced by other climatic parameters. Hourly precipitation analysis in Côte d’Ivoire over a 10-year period showed a negative trend in extreme hourly rainfall. Studies linking floods with atmospheric circulation patterns indicate that flood-producing precipitation in different geographic regions is associated with distinct synoptic systems, many of which have now been identified [Soro et al., 2016].
Investigation of extreme rainfall in China using satellite imagery and multiple Doppler weather radars revealed that catastrophic heavy rainfall resulted from a mid-level convective system around 14 km altitude. Local convergence driven by regional topography and the intrusion of a cold tongue at the mountain base intensified the rainfall system [Li et al., 2020]. Additional analyses using principal component methods identified six major weather patterns and three key centers of heavy rainfall. These studies also demonstrated that the spatial occurrence of large-scale heavy rainfall events was closely related to urban density, indicating that urbanization significantly influences the magnitude, intensity, and spatial concentration of such events [Li et al., 2021].
The objective of the present study was to extract the key dates associated with precipitation variability, not only the extreme wet and dry events but also the spatial extent involved, and to identify temporal changes in precipitation throughout the study period on a month-by-month basis.
Methodology
This study employed daily precipitation data from 117 synoptic stations across Iran with a shared statistical baseline for the period 1986–2020. These data were obtained from the Data Provision Department of the Iran Meteorological Organization. Because most precipitation in Iran occurs during autumn, winter, and spring, the analysis focused on the period from October to June of each year.
Homogeneity testing was performed using XLSTAT (Standard Normal Homogeneity Test, SNHT), an add-in for Microsoft Excel. Stations with the longest common statistical period and sufficient data continuity were selected. Prior to 1986, the number of synoptic stations was too limited to allow reliable spatial interpolation; and although the number of stations increased after this date, the available record still does not constitute a long-term climatological baseline. A larger number of stations with balanced spatial distribution leads to more accurate interpolation and more reliable assessment of spatial and temporal precipitation variations. Ultimately, among the available stations, those with more complete and reliable meteorological records were selected for analysis. No data reconstruction was applied in this study. Despite the increase in the number of synoptic stations in recent years, their spatial distribution remains heterogeneous and not uniformly distributed across the country. Monthly mean precipitation for each station was calculated using the arithmetic mean. To determine whether each month at each station was wetter or drier than average, the precipitation of each month was compared with the long-term monthly mean. The resulting values were categorized as positive, zero, or negative. A positive value indicates precipitation above the long-term mean; zero indicates equality with the mean; and a negative value indicates lower-than-average precipitation.
The deviations for all stations were calculated using MATLAB 2013, and the interpolated spatial patterns were generated using SURFER 24.3 to identify areas experiencing above-average or below-average precipitation, as well as the magnitude of these deviations.
A dual-line chart was constructed to illustrate the number of wet stations (precipitation above the mean) and dry stations (precipitation below the mean) in each month. The inverse relationship between these two categories, where an increase in one corresponds to a decrease in the other, helps identify the overall spatial patterns of wetter or drier conditions during the study period. The distance between the two lines and their points of intersection provide insight into the dominance and temporal shifts of these precipitation regimes.
Findings
Mean Precipitation and Variability of Precipitation Zones
The mean precipitation from October to June during the period 1986–2020 and the variations in precipitation zones were examined.
October: The highest precipitation levels in this month occurred in the northern regions, reaching up to 260 mm, followed by the northwest and the Kouhrang area of Isfahan with more than 120 mm. The southernmost boundary of precipitation above 20 mm extended to the southern parts of Lorestan Province. However, a vast portion of the country received less than 20 mm of rainfall. Two seven-year dry periods were observed during 1994–1998 and 2001–2008. Only in six years did the number of wet stations exceed that of dry ones. In recent years, the number of wet Octobers has increased slightly.
November: The area receiving more than 20 mm of precipitation expanded compared to October and reached the southwestern regions. The northern parts of the country still recorded the highest precipitation, around 260 mm. Kouhrang, with more than 120 mm of rainfall, had the highest precipitation after the Caspian Sea coasts. Western, northwestern, and northeastern regions also recorded more than 80 mm. At the beginning of the study period, an eight-year dry spell was evident. In eleven years, the number of wet stations exceeded dry ones, mostly in the later years.
December: In this month, the western, southwestern, and northern regions received comparatively more rainfall. Although the spatial distribution of precipitation showed little change compared to November, the amount recorded at stations generally increased. Southern regions, parts of the central plateau, and the northeast received more than 50 mm, whereas the east, southeast, and central parts had lower averages. A wet period was observed from 2000 to 2004. In ten years, the number of wet stations exceeded the dry ones. Unlike October and November, dry zones increased in recent years.
January: Rainfall in January expanded further eastward compared to December. The western to northwestern regions received higher rainfall, and precipitation extended farther south. The lowest precipitation occurred in central and eastern regions, with the highest levels in the southwest, west, and north. Several short and long dry periods appeared in recent years, and in twelve years, dry zones exceeded wet ones.
February: The maximum precipitation increased from 180 mm in January to 200 mm, mainly in western regions and the western Caspian coast. Rainfall decreased in the east and increased in central areas. Other stations experienced little change. At the beginning of the study period, wet and dry areas were similar in extent, but in recent years dry zones surpassed wet ones in both number and spatial coverage. Only in six years did wet stations outnumber dry ones.
March: Peak precipitation shifted from the Caspian coastal regions to the southwestern heights and Kouhrang. Areas receiving less than 30 mm expanded. Most stations recorded lower precipitation except for limited regions such as the central Zagros elevations, where precipitation increased. Patterns of dry and wet periods were clearly visible, and recent years did not differ significantly from earlier periods. In seven years, above-average precipitation covered a larger area of the country.
April: Rainfall patterns changed noticeably compared to March. Peak precipitation remained in Kouhrang but decreased by nearly 100 mm. Central and southeastern regions recorded almost no precipitation, whereas northwest, west, and parts of the southwest received up to 70 mm. Large central areas recorded up to 20 mm. An eight-year dry spell occurred from 2000 to 2008. In twelve years, above-average rainfall covered extensive areas, with broad and heavy precipitation reported in recent years.
May: Peak precipitation occurred in the northwest, reaching up to 55 mm. The lowest precipitation (up to 5 mm) was observed in central, southern, and southeastern regions. Only in four years did wet zones exceed dry ones, and widespread dryness dominated recent years.
June: Rainfall declined sharply compared to May. The highest precipitation occurred in the northern regions but did not exceed 46 mm, although these regions were the wettest relative to the rest of the country. Southern Sistan and Baluchestan saw a slight increase and was the only region with higher rainfall than in the previous month. Most of the country received less than 4 mm. Only in three years did wet stations slightly outnumber dry ones. In recent years, a pronounced dry period has developed, with dry zones far more extensive than wet ones.
Wettest and Driest Months Based on Spatial Extent
The wettest October, driest November, and driest January during the 35-year period all occurred in 1987, indicating that this year experienced the most expansive wet and dry zones. January and February of 1996 had the largest areas with above-average precipitation. In March and May 2008, the widest wet zones were recorded.
Absolute dryness with below-average rainfall across all 117 stations occurred in November and January of 1987. The highest number of stations with above-average rainfall, 102 stations, was recorded in March 1996. December and January had the broadest spread of above-average precipitation; in 15 years, more than half of the stations experienced above-average rainfall. In June, only three years had more than half of the stations recording above-average precipitation, while in the remaining years, most stations had below-average rainfall.
Discussion
In addition to the statistical analysis of precipitation, the objective of this study was to determine the spatial extent of areas experiencing below- or above-average precipitation from October to June. The spatial coverage of above-average precipitation also reflects the extent of the precipitation-inducing systems. Identifying the years with the widest precipitation coverage for each month provides a basis for examining the synoptic systems responsible for such events.
Assessing monthly precipitation anomalies relative to the same month in other years is important because, when analyzing the temporal pattern of precipitation across consecutive months, climatic factors such as solar angle remain constant. However, large-scale systems, whose strength depends on the solar angle, do not behave uniformly across the months of a single year. Thus, comparing each month to its own long-term average while holding constant the factors that do not change allows us to identify additional causes that make each month unique relative to its climatological norm. In this way, each month’s precipitation is compared only with its own type, and not with months that do not share similar atmospheric conditions.
Using the dual-line diagram applied in this study, the spatial extent of heavy precipitation for each month can be clearly observed and quantified, along with the magnitude of precipitation deficits. Moreover, changes in the spatial coverage of heavy precipitation from year to year for each month are identifiable. The strength and expansion of above-average precipitation vary noticeably from one month to another. For example, in autumn the spatial extent of above-average precipitation is smaller than in winter. Additionally, different regions of the country enter the category of high-precipitation zones in different months; for instance, the eastern and southeastern regions tend to be included in the wet zone at times when most other regions fall outside it. Overall, analysis of the dual-line diagrams provides extensive information about the precipitation characteristics of each month. Identifying the wettest and driest years, along with the spatial extent of these conditions for each month during the study period, is one of the distinguishing features of this research.
Previous studies have examined precipitation trends and variability across different regions of the country using the Mann–Kendall test. This test has been applied to evaluate the significance of precipitation trends, spatial patterns of rainfall amounts, the highest and lowest coefficients of variation, and the frequency of dry and wet periods in the study regions [Mirian et al., 2022; Mazidi et al., 2022; Amani et al., 2020]. Markov chains have also been widely used as a stochastic model for describing sequences of probabilistic events in which the probability of each event depends solely on the previous state. Numerous studies have used this approach to analyze precipitation trends and ultimately forecast rainfall. Analyses have included the persistence and frequency of wet and dry days, the probability of rainfall occurrence, rainfall amounts, and return periods. Time-series trends of wet and dry periods, their intensities, and the probability of transitions between wet and dry states have also been investigated using this method [Ghaznavi et al., 2021; Iranpour et al., 2019].
Drought monitoring can be carried out from climatological, hydrological, agricultural, economic, and social perspectives. In climatological drought assessments, the Standardized Precipitation Index (SPI) is widely used. The SPI at 12-, 24-, and 48-month timescales has been calculated for selected stations, and interpolated maps have been produced. Moran’s spatial autocorrelation statistic and hot-spot analysis have been used to identify drought patterns and map their distribution. Findings for Fars Province indicate that nearly all parts of the province experience drought with varying severities. Results also show that as the temporal scale increases, SPI values rise, while the absolute magnitude of drought decreases [Ghanbari, 2020]. The SPI is also used to identify wet and dry conditions in different regions and is extensively applied for drought monitoring and forecasting. In this method, wetness and dryness levels are classified into grades with defined thresholds [Bazrafshan et al., 2011; Shokohi et al., 2018; Mohtashami et al., 2015].
Changes in the waiting time between rainfall events have also been studied. Using geostatistical kriging, daily precipitation values were interpolated on 6×6 km grid cells. A non-parametric Mann–Kendall test at the 90% confidence level was applied to the time series of waiting times between precipitation events, and rate of change was estimated using linear least squares. Findings indicated that the waiting time between rainfall events has increased during the cold months of the year, whereas the intervals between consecutive rainfall events have shortened during the warm months [Darand et al., 2019].
Monthly rainfall behavior at spatial and temporal scales has also been examined through the relationship between precipitation and topographic parameters. Although a statistically significant relationship exists between monthly rainfall variability and spatial–topographic factors, correlations have not been particularly strong. Using cluster analysis, five regions were identified. These regions generally extend northward from latitude 32°N, while the area south of 32°N tends to follow longitudinal patterns [Asakare et al., 2021]. Heavy precipitation has also been analyzed using a threshold of the 95th percentile and non-parametric tests such as Mann–Kendall, Sen’s estimator, and simple linear regression. Results indicated a significant increasing trend only at the Astara station, while other stations showed non-significant upward trends [Zeynali et al., 2020].
In the present study, only one threshold, the long-term mean, was used to classify each station into wet (above-average precipitation) or dry (below-average precipitation). This binary classification facilitates the analysis of related parameters, such as the duration of wet and dry periods, the spatial differences between wet and dry areas, and their temporal variability. Using multiple thresholds would complicate the classification process and make it more difficult to analyze the relationships among categories. In the Mann–Kendall test and Markov chain analysis, if precipitation deviates substantially from normal conditions, trend significance may no longer be detectable, which renders forecasting with these methods ineffective. Given the influence of climate change and global warming, an increase in non-normal precipitation events is expected. Another consequence of climate change is the spatial–temporal displacement of precipitation patterns, a highly parameter and chaotic atmospheric parameter. While previous studies have assessed precipitation intervals in warm and cold seasons using linear least squares, the present study identifies intervals of wet and dry conditions for each month across consecutive years. This study does not aim to analyze precipitation trends or provide forecasts; rather, it focuses on extracting key spatiotemporal variations in precipitation to facilitate the investigation of their underlying causes.
One of the limitations of this study is its use of a single threshold, which does not allow for grading the severity of wetness or dryness in the study areas. Future research may incorporate multiple precipitation thresholds appropriate to regional climates if the research objectives require such refinement. Finally, given the identification of the driest and wettest months, as well as the years with the widest spatial extent of dryness and heavy precipitation, it is recommended that future studies complement these results with synoptic and dynamical analyses of the responsible atmospheric systems.
Conclusion
Precipitation in Iran generally occurs over relatively limited spatial extents. Although below-average precipitation is observed across all stations at times, there has never been a case in which all stations across the country experience above-average precipitation simultaneously.
Acknowledgments: We sincerely thank the Iran Meteorological Organization for providing the data used in this study.
Ethical Permission: No ethical issues were applicable.
Conflict of Interest: The authors declare no conflicts of interest.
Author Contributions: Derikvand S (First Author), Main Researcher (50%); Nasiri B (Second Author), Discussion Writer (15%); Ghaemi H (Third Author), Methodologist (15%); Karampoor M (Fourth Author), Data Analyst (10%); Moradi M (Fifth Author) – Introduction Writer (10%)
Funding: This study is derived from the first author’s doctoral dissertation titled “Investigation of the Impact of the Overlying Atmosphere on the Lower Atmosphere in Intensifying or Weakening Iran’s Precipitation during the Cold Season”, conducted at the University of Lorestan under the supervision of the second and third authors and with consultation from the fourth and fifth authors.
References
1. Asakare H (2007). Temporal-spatial changes of rainfall in Iran during recent decades. Tehran. Journal of Geography and Development. 5(10):145-164. [Persian] [Link]
2. Asakereh H, Masoodian S A, Tarkarani F (2021). Spatial Analysis of variation of inter-annual distribution of precipitation in Iran over recent decades (1970-2016). Journal of Spatial Analysis Environmental Hazards. 8(4):103-122. [Persian] [Link]
3. Amani M, Zohorian M, Borna R (2021). Spatial analysis of drought trend and calculation of reliable rainfall in Khuzestan province. Quarterly Geography and Regional Planning.1(13):97-109. [Persian] [Link]
4. Bazarafshan U, Mohseni Saravi M, Malekian A, Moeini A (2011). Investigating the drought situation of Golestan province using Standard Precipitation Index (SPI). Iranian Journal of Rangeland and Desert Research. 18(44):395-407. [Persian] [Link]
5. Darand B, Ebrahimi M (2019). Tempo-spatial analysis of precipitation waiting time duration changes in Kurdistan province. Marvdasht. Water Resources Engineering. 12(41):17-30. [Persian] [Link]
6. Davarpanah Gh (2004). The effects of flood and drought and providing solutions to deal with them. Isfahan. Journal of Water and Wastewater. 15(1):73-76. [Persian] [Link]
7. Doostan R (2020). An analysis of rainfall changes in Iran. Journal of Climate Research. 10(40):13-25. [Persian] [Link]
8. Farajzadeh M (2012). Climatic hazards of Iran. 1st ed. Tehran: Publications of Humanities Research and Development Center( SMT). [Persian] [Link]
9. Ghanbari A (2020). Spatial Analysis of Fars Province Climatic Drought from the Period 1990 to 2014. Nivar. 44(110):53-64. [Persian] [Link]
10. Ghaznavi M, Mosaedi M, Qabaei Souq (2021). The effect of breaking the time series of rainfall data on changes in drought characteristics (A case study of the cities of Tabriz and Arak). Journal of Water and Soil. 35(3):409-425. [Persian] [Link]
11. Hasanvand Z, Yarahmadi D, Lashkari H, Mirhashemi H (2022). Fractal analysis of daily rainfall in Karkheh and Dez watersheds. Larestan. Natural Geography Quarterly. 15(55):115-129. [Persian] [Link]
12. Heydari A, Shahkoui I (2017). Investigating the probability of occurrence of dry and wet periods in Mazandaran using Markov chain. Environmental studies( Natural resources and sustainable development). 4(36): 16-31. [Persian] [Link]
13. Iranpour F, Zohrevandi H (2019). Analysis of daily rainfall in Tabriz to investigate the probability of the continuation of dry and wet days. Climatology Research. 9(36): 91-105. [Persian] [Link]
14. Jahanbakhsh Asl S, Sari Saraf B, Asakareh H, Shirmohammadi S (2022). Identifying extreme precipitation events in western Iran (1965-2016). Journal of Spatial Analysis Environmental Hazards. 26(79):115-125. [Persian] [Link]
15. Khosravi A, Azari M (2022). Determining the temporal and spatial trend and point of change of temperature and precipitation in Kashf River basin. Scientific Journals Management System. 22(66):289-306. [Persian] [Link] [DOI:10.52547/jgs.22.66.289]
16. Li S, Li G, Wang X, Li C, Liu H, Li G (2020). Precipitation Characteristics of an Abrupt Heavy Rainfall Event over the Complex Terrain of Southwest China Observed by the FY-4A Satellite and Doppler Weather Radar. Water. 12(9):2502. [Link] [DOI:10.3390/w12092502]
17. Li W, Bi X, Sheng L, Luo Y (2021). Sun Jianhua modulations of synoptic weather patterns on warm-sector heavy rainfall in south China: Insights from high-density observations with principal component analysis. Frontiers in Earth Science. (9):1-12. [Link] [DOI:10.3389/feart.2021.678230]
18. Masoudian A, Kaviani M (2007). Climatology of Iran. Isfahan: Isfahan University Press. [Link]
19. Mazidi A, Salamati Hormozi V, Omidvar K, Mozafari G (2022). Revealing the temporal-spatial changes of precipitation in southeast Iran. Geographical explorations of desert areas. 5(18):47-64. [Persian] [Link]
20. Mohammadi B, Masoudian A (2010). Synopsis analysis of heavy rainfall in Iran, a case study. November 1373. Geography and Development. 8(19):47-70. [Persian] [Link]
21. Mohammadyarian M, Taavosi T, Khosravi M, Hamidianpour M (2019). Zoning of Iran's extreme rainfall regime in the 20-year period. Tehran. Journal of Geographical Research. 34:192-183. [Persian] [Link] [DOI:10.29252/geores.34.2.183]
22. Mohteshmi A, Nasiriyan A (2015). Drought monitoring of South Khorasan province using standard precipitation index (SPI). Iran National Irrigation and Drainage Congress. 17(1):177-181. [Persian] [Link]
23. Mirian M, Moradi M, Ghaemi H, Nasiri B, Karampoor M (2022). Analysis of statistics and anomalies of 50-year rainfall of Iran's synoptic stations (period 1960-2010). Applied Research of Geographical Sciences. 67(17):327-343. [Persian] [Link]
24. Sadaqat M (2001). Iran's water resources and issues. 7th ed. Tehran: Payam-e Noor University Publications. [Persian] [Link]
25. Salimi Mostala A, Khoshkhoo Y, Gholizadeh M (2019). Assessing and zoning the risk of extreme rainfall in western Iran. Journal of Water and Soil Consevation. 26(1): 49-69. [Persian] [Link]
26. Shokohi A, Hosseini N, Bakhtiari A (2018). Flood forecasting using the Standard Precipitation Index (SPI) on a daily scale.Tabriz. Journal of Civil and Environmental Engineering (University of Tabriz). 48(91):27-35. [Persian] [Link]
27. Soro, GE, Noufé D, Goula Bi TA, Shorohou B (2016). Trend analysis for extreme rainfall at sub-daily and daily timescales in Côte d'Ivoire. Climate. 4(3):37. [Link] [DOI:10.3390/cli4030037]
28. Torabipoodeh H, Izadjo F, Hamehzadeh P (2018). Analysis of changes in total precipitation and effective precipitation in the whole of Iran. Iranian Water Research Journal. 12(4):1-10. [Persian] [Link]
29. Zeynali BA, Minaei S, Foroutan M (2020). Statistical and synoptic study of super heavy rainfall in Guilan province. Jornal of Environmental Sciences Studies. 7(4):5682-5695. [Persian] [Link]