Climatology of the Sistan Levar wind: Atmospheric dynamics driving its onset, duration and withdrawal

Highlights

• Analysis of Levar onset, withdrawal and duration for a period of about 40 years.

• Typical duration of Levar wind is about 135 days (mid May to early October).

• The Caspian-Sea high strongly modulates the Levar onset, duration and intensity.

• The Indian/Pakistan thermal low is a regulatory factor for Levar onset and withdrawal.

• Connection between Levar and CasHKI via changes in sea-level pressure over the Caspian Sea.

Abstract

Levar wind is the dominant meteorological and climatic feature in east Iran, blowing from Central Asia to the northern coast of the Arabian Sea. It is also known as “120-days wind” due to its mean duration in the summer season. Although violent and responsible for massive dust storms in southwest Asia, long-term climatology and atmospheric dynamics that facilitate genesis and dissipation of Levar have not been well documented. This study uses the two-phase regression method for determining the onset, dissipation and duration of Levar during a 41-years period (1972–2012), based on identification of the change points in wind data series at Zabol meteorological station. The mean duration of the Levar period is estimated to 135 ± 24 days, with a remarkable inter-annual variability. The mean Levar onset is determined around 22 May ±23 days, while the withdrawal on 8 October ±25 days. On average, the onset and dissipation of Levar occur between wind speeds of 4.6–7.7 ms⁻¹ in spring and 3.9–6.7 ms⁻¹ in autumn, respectively. A comprehensive analysis is performed for the first time to examine meteorological dynamics that are associated with onset, duration and withdrawal of Levar. The early onset years are characterized by stronger winds during May, while no considerable changes in wind are detected between early and late withdrawal years. Changes in the mean sea-level pressure (MSLP) dipole between the Caspian Sea and India/Pakistan forces the onset of Levar. Therefore, early (late) onset (withdrawal) of Levar are driven by changes in MSLP over these areas, while local dynamics and topography also play an important role. Levar seems to be modulated by CasHKI (Caspian Sea Hindu Kush Index), while changes in the Caspian-Sea High (majorly) and in the Indian/Pakistan monsoon thermal low (secondarily) affect the Levar intensity.

Introduction

The Sistan basin or watershed (~15,000 km²) is a topographic-low area in the borders of Iran and Afghanistan, well known for its aridity, massive dust storms and, particularly, for its violent seasonal wind, the so called Levar. Sistan wind (Levar) is the main climatic feature in east Iran, southwest and Central Asia that is blowing southeastward with high intensity during the period of May to September (Middleton, 1986; Alizadeh Choobari et al., 2013, Alizadeh Choobari et al., 2014; Abbasi et al., 2019). The wind is responsible for high erosion of the dried playas in the Hamoun ephemeral lakes (Abbasi et al., 2018; Behrooz et al., 2019), while its intensity usually overcomes 20–25 ms⁻¹ (Kaskaoutis et al., 2015a; Abbasi et al., 2018; Behrooz et al., 2017a), thus causing severe dust storms (Kaskaoutis et al., 2015b; Rashki et al., 2019; Karami et al., 2021). These interrelated phenomena i.e. Hamoun dried lakes, Levar wind, Aeolian erosion, wind channeling effect through the mountains, render the Sistan basin as one of the windiest and dustiest arid lands in southwest Asia and over the globe (Goudie, 2018; Gholami et al., 2020a; Rashki et al., 2021).

The first study about the Sistan wind is from Huntington (1905) during his field visit in Sistan in 1904. He described the wind as “…During three months of the summer, a remarkably constant wind blows night and day with great violence from the north-northwest. It seems to be the northward continuation of the trade-winds deflected to the west by the prevailing trend of the mountains…”. At the same period, McMahan, 1906a, McMahan, 1906b, after his field visit, estimated the wind's duration to about 120 days, stating that the wind was blowing without interruption from about May to early September. The persistence and violence of Levar have been highlighted in several recent works based on measurements at meteorological stations in Sistan and surrounding areas, reporting highest wind speeds of about 20–30 ms⁻¹ (~80–110 km per hour) (Kaskaoutis et al., 2015a; Behrooz et al., 2017a; Abbasi et al., 2018).

Previous studies have attributed the genesis of Levar to pressure gradients between synoptic, regional and local atmospheric circulation systems over Central and southwest Asia (Hosseinzadeh, 1997; Saligheh, 2010; Gandomkar, 2010; Khosravi, 2008, Khosravi, 2010; Hamidianpour et al., 2016), but without detailed investigation. Such systems include the Indo-Pakistan thermal low that initiates the southwest summer monsoon, the high-pressure conditions over the Caspian Sea expanded over Turkmenistan (Mofidi, 2007; Hamzeh et al., 2016), local systems of high pressure over South Khorasan mountains in east Iran and thermal lows developed within the Sistan basin (Hamidian Pour, 2014, Hamidian Pour et al., 2016; Saeedi and Khoshakhlagh, 2019). Therefore, the Sistan wind is a climatic characteristic and a repeated phenomenon occurring during a specific period and then dissipates with the elimination of its driving forces. Alizadeh Choobari et al. (2014) reported that Levar is related with a high-pressure system in the mountains of Pamir and with the monsoon thermal low, while more recent works showed that Levar is modulated by changes in pressure gradient between the Caspian Sea and the Hindu Kush – Pamir mountains (Kaskaoutis et al., 2017, Kaskaoutis et al., 2018a, Kaskaoutis et al., 2018b). The gradient of the pressure anomalies between these regions was quantified via the Caspian Sea – Hindu Kush Index (CasHKI), which was found to modulate the dust activity, loess deposits and climate in southwest and Central Asia (Kaskaoutis et al., 2016; Li et al., 2020a; Lu et al., 2020). Previous works also characterized Levar as a return flow of the Indian summer monsoon, but till now its long-term climatology and the dynamic factors that are associated with its genesis and dissipation are poorly understood. Apart from Levar and the dominant southwest summer monsoon flow over the Arabian Sea (Suresh et al., 2021), other seasonal wind patterns in southwest Asia, Middle East and east Mediterranean, with distinct genesis and dissipation driving factors in summer, are Shamal over the Mesopotamia and east Arabia (Yu et al., 2015; Francis et al., 2017) and Etesians in the Aegean Sea (Metaxas and Bartzokas, 1994; Nastos et al., 2002; Poupkou et al., 2011), which also modulate the dust aerosol and pollution transport over these regions (Lelieveld et al., 2002; Sciare et al., 2008; Mashat et al., 2019; Mohammadpour et al., 2020).

The change point is a ubiquitous feature in climatic data series (Lund and Reeves, 2002), attributed to several reasons like changes in recording instruments, station locations, atmospheric dynamics, etc. For climatic time series, an increasing or decreasing trend is affected by changes in large- and small-scale atmospheric dynamics, but change points can modulate the significance of the trends (Kambezidis et al., 2020) and determine the onset or dissipation of specific phenomena, like frontal systems, dust storms, etc. (Maghrabi et al., 2011; Hamzeh et al., 2021; Huo et al., 2021). Change points are applicable for several phenomena that could affect the weather and climate dynamics in the Middle East like the advent of sub-tropical high pressure during the warm season (Gholipour et al., 2017) or the Arabian high pressure during the fall season (Lashkari and Mohammadi, 2018), for the initiation of prolonged dry periods (Kaskaoutis et al., 2012; Notaro et al., 2015), for the onset, duration and breaks of Shamal wind (Yu et al., 2016) and summer-monsoon flow (Liu et al., 2013; Manoj et al., 2011) and for the expansion of the Sudanese low over southwest Iran (Mohammadi et al., 2012).

Therefore, statistical methods for detecting the change-point can be very effective for studying the timing of onset, termination or duration of such phenomena, which are especially important for climatological studies (Choudhury et al., 2019; Mohammadpour et al., 2020; Li et al., 2021). There are several statistical methods for detecting the change point in a data series like the Pettitt test (Pettitt, 1979), the Buishand test (Buishand, 1982), the standardized homogeneity test (Alexanderson, 1984, Alexanderson, 1986) and the two-step regression (Solow, 1987), among others. These methods have been used in numerous studies and for a variety of topics, such as determining the points of change in rainfall (Thompson, 1984), homogenizing meteorological time series (Gullet et al., 1991; Kambezidis et al., 2020) and identifying climate change (Solow, 1987; Lund and Reeves, 2002). In addition, Vincent (1998) used the two-phase regression method for the identification of inhomogeneities in Canadian temperature series and the results showed mean shifts in observed temperatures that were attributed to change points. Cook and Buckley (2009) applied the two-phase regression method on daily precipitation data for identifying the onset, withdrawal and duration of the monsoon season in Thailand. In this study, we used the two-phase regression method (Solow, 1987; Lund and Reeves, 2002) for identifying the change points in annual Levar wind data series aiming to determine the onset, withdrawal and duration of Levar.

Few previous works have attempted to assess the onset and withdrawal of the Sistan wind using different methods (Mofidi et al., 2014; Masoodian, 2014), but long-term climatology of Levar, the local and synoptic meteorological conditions that are associated with its genesis and dissipation are poorly understood. The main objective of the present study is to determine the onset, withdrawal and duration of the Levar wind through the two-phase regression method, covering a long-term (41 years) period from 1972 to 2012. The timing of the wind onset and withdrawal may constitute a great tool for weather forecasters and policy makers in warning the local population about the devastating sand/dust storms associated with severe Levar winds. In addition, this study examines the synoptic meteorological patterns that are associated with early or late onset (withdrawal) of the Levar period, focusing on the dynamic role and the relative influence of the prevailing meteorological/climatic features in Central and southwest Asia like the Caspian-Sea High and the summer-monsoon low. The results of this study are especially important for atmospheric scientists due to heavy dust storms associated with intense Levar, and engineers regarding the performance of photovoltaic plants and wind parks for renewable energy resources.