3-1- Climatic study of tropical cyclones in the Arabian Sea
The statistical analysis of monthly TC frequencies in the AS for the period 1980–2023 indicates a concentration of storms in the months of June, May, October, November, and December (Fig. 2). TC occurrences were categorized into pre-monsoon and post-monsoon seasons, with May and June representing the pre-monsoon period, and October, November, and December corresponding to the post-monsoon season (Evan et al., 2011; Al-Maskari, 2012). Notably, no storms were recorded in January, February, and April, while only one case was documented in March within the AS.
The examination of cyclones in the AS in relation to teleconnection indices (Fig. 3) reveals that a majority of TCs occurred during a negative phase (La Niña) (<-0.5) or a neutral phase (index between 0.5 and − 0.5) of the ENSO. Specifically, 36.56% of cyclones were observed during the La Niña phase, 38.71% during the neutral phase, and 24.73% during the El Niño phase. Furthermore, these TCs predominantly formed during a neutral phase of the IOD (index between 0.4 and − 0.4), with 70.9% occurring during a neutral phase, 9.78% during a negative phase, and 14.13% during a positive phase. Notable instances include TC Mekunu during La Niña and a neutral IOD, Shaheen during severe La Niña and a neutral IOD, and Biparjoy during a neutral phase of ENSO and a neutral IOD.
Several studies have highlighted the warming of eastern North Indian Ocean (NIO) waters as a key factor in storm formation in the NIO region (Bay of Bengal and AS), particularly during the La Niña phase of ENSO (Leung and Leung, 2002; Lin et al., 2020). Additionally, research suggests a higher occurrence of TCs during the negative or neutral phase of ENSO and the positive phase of IOD (Mahala et al., 2015). In Fig. 3, the La Niña phase is represented by large circles, the neutral phase by medium circles, and the El Niño phase by smaller circles. Similarly, the negative phase of IOD is denoted by large triangles, the neutral phase by medium triangles, and the positive phase by smaller triangles.
Figure S1 illustrates the origin points and trajectories of TCs in the Arabian Sea during the pre-monsoon months (May and June) and post-monsoon months (October, November, and December). In May, cyclones predominantly initiated in the east, with some also originating in the south and southwest of the AS. Those forming in the eastern region generally followed a north-eastern trajectory, making landfall in Gujarat (India) and the southeast of Pakistan. TCs originating in the south weakened after landfall over Oman and Yemen to the west, while those forming in the southwest moved westward, entering the Gulf of Aden and Somalia. In some instances, they directly made landfall on the Horn of Africa (Somalia) (Figure S1-a). During June, cyclones formed in the east and center of the AS. TCs originating in the southeast typically moved north and northeast, weakening upon landfall in India. Cyclones forming in the central and eastern parts moved westward, with some reaching Oman and others decaying in the western part of the sea. In rare cases, they reached the Oman Sea and the coasts of Iran (Figure S1-b).
In the post-monsoon season in October, cyclones formed in the east and south of the sea. Those originated in the east continue north and northeast, making landfall over Gujarat (India). Cyclones from the south moved westward, landing in Oman, Yemen, or the Horn of Africa (Somalia) (Figure S1-c). In November, more cyclones formed in the south and moved westward, making landfall over Somalia via the Horn of Africa and the Gulf of Aden. In rare cases, they might be landed in Yemen or decayed in the western part of the AS. Additionally, a small number of cyclones moved north and then northeast, landing on India. In November, cyclones formed below 10°N, moving westward towards the Horn of Africa (Figure S1-d). In December, similar to November, cyclones formed in the south of the AS, with more below 10°N. Most cyclones moved westward, landing on the Horn of Africa or the Gulf of Aden. Cyclones originating from the southern coast of India or the coast of Sri Lanka moved northwest, changing direction towards the northeast before making landfall in India (Figure S1-e).
3-2-Investigating the path of TCs: Biparjoi, Shaheen and Mekunu
Analysis of the trajectories of TCs Biparjoy, Shaheen, and Mekunu is presented in Fig. 4. Storm Shaheen originated along the eastern coast of the Bay of Bengal and, after traversing the Indian subcontinent, entered the AS. Subsequently, it initially approached the northern shores of the Oman Sea located in the southeast of Iran. However, after a change in direction, it made landfall to the north of Oman. It is important to note that, in this study, the path of Storm Shaheen is exclusively investigated and simulated within the AS.
The trajectory of TC Mekunu indicates its origin in the southwest of the AS, moving northwest. Eventually, Mekunu made landfall in the south of Oman, close to the border with Yemen.
TC Biparjoy developed in the central AS, moving initially northward and then northeastward. It eventually made landfall on the northwestern coast of India, close to the border with Pakistan.
3-3-Tropical cyclone Biparjoy
3-3-1-Satellite data and products
Figure 5 displays true color images of MODIS/TERRA on June 6 and 16, 2023. On June 6, the cloud mass associated with TC Biparjoy was situated to the south of the AS. By June 16th, Biparjoy made landfall, covering portions of Gujarat (India) and certain coastal areas in the south of Pakistan
Figure 6 depicts the effective radius and Cloud Optical Depth (COD) as captured by MODIS/TERRA on June 6, 2023. In a significant portion of the southern AS, where TC Biparjoy was present, clouds in the ice phase are observed. Around the cyclone's center, the effective radius of cloud in the ice phase measured approximately 30µm. The maximum effective radius, around 60µm, is observed in the southwest of the storm's eye. Water phase clouds, characterized by an effective radius between 15 and 30µm, were present farther from the center of Biparjoy (Fig. 6-a). Near the eye of the storm, the COD in the ice phase was higher, reaching approximately 100. Notably, in areas where the effective radius of the cloud was larger, the optical depth was generally lower (Fig. 6-b).
The CALIPSO satellite products on June 9, 2023, at approximately 10 UTC, as it traversed the AS are presented in Fig. 7. Notably, CALIPSO version 4.51 products were unavailable on this date, so version 3.41 was utilized, where the cloud type is not specified. Total Attenuated Backscatter values between latitudes 12 to 18 °N indicate a significant decrease in altitude, ranging from about 7 to 17 meters, attributed to the presence of clouds. In the latitude of 10 °N (Fig. 7-b), tropopause folding and a relative decrease in potential temperature are observable.
CALIPSO cloud phase product reveals that the clouds in this region were predominantly in the ice phase. However, water phase clouds are also observed in certain areas at lower altitudes within the troposphere (Fig. 7-c). Higher latitudes exhibited notable concentrations of dust particles up to an altitude of about 5 km, potentially influencing cloudiness, cloud microphysics, precipitation, and even the intensity and trajectory of TCs (Fig. 7-d). It's essential to note that a comprehensive examination of these phenomena falls beyond the scope of this article and necessitates further scientific and detailed investigations.
3-3-2-Atmospheric anomaly patterns
The MSLP anomaly reveals a significant pressure drop (-40hPa) in the northern part of the eastern AS between 65 to 70 °E and 12 to 25 °N. Additional pressure drops are observed over the Bay of Bengal and Turkmenistan. Conversely, pressure increases are evident in northwest India, east of Pakistan, and notably over the Himalayas (+ 60hPa) (Fig. 8-a).Examining the 30-year (1993–2022) long-term pattern, a band of low pressure is consistently observed over the east and north of India, Pakistan, southern Afghanistan, southern Iran, the eastern Arabian Peninsula, the Persian Gulf, and the Oman Sea (Fig. 8-b). However, during the days of TC Biparjoy, the low-pressure system extended towards the AS, particularly to its east. A 996hPa low-pressure area formed over the east of the AS, indicating the strengthening of the low-pressure system in this region (Fig. 8-c).
One of the influential factors affecting the trajectory of a TC is the flow direction at different altitudes of the atmosphere, (Deng et al., 2010; Tang and Chan, 2016). The 850hPa wind anomaly pattern reveals a robust anomaly around 70 °E in the east, directing toward the western coasts of India and southern Pakistan. This anomaly indicates a weakening of the westerly winds and a strengthening of the southerly/southwesterly winds from the southeastern AS through the northeast (Fig. 9-a). The long-term pattern of 850hPa wind depicts the Somali low-level jet stream (LLJ) extending from the Horn of Africa to the central parts of the AS and the southwest coast of India, with a jet core speed exceeding 21 m/s. In the east and northeast of Iran and the west of Afghanistan, the winds were the northeasterly and northerly, reaching a maximum speed of 15 m/s. Additionally, northwesterly and northerly flow is observed from Iraq and the Persian Gulf toward Arabia (Fig. 9-b). During TC Biparjoy, the 850hPa wind direction in the left exit region of the LLJ shifted, flowing southwest-northeast towards western India and southeast Pakistan. A cyclonic circulation formed in this region, supported by strengthened northerly winds from the west of Afghanistan and east of Iran. These conditions contributed to increased convergence and moisture transfer to the storm region. Furthermore, an anticyclone over Arabia is observed with a relative strengthening of wind speeds. Additionally, an increase in westerly winds is noted over the northeastern part of India compared to the long-term average (Fig. 9-c).
The 500hPa potential vorticity anomaly in Fig. 10-a reveals a positive anomaly at 65–70 °E, with the maximum value (+ 1.6) occurring at 21 °N. Concurrently, the investigation of SST during TC Biparjoy in Fig. 10-b indicates an increase of 1 to 2°C in the AS, with a higher increase of 3°C in the west and 2°C in the east. Similar increases in SST are observed in the Red Sea and the Oman Sea. Increasing of the SST and transferring heat and moisture into the Cyclone will strengthen the storm (Asadi et al., 2021). Warmer waters provide the necessary energy to enhance storms and can also influence their trajectory by creating favorable conditions (Trenberth, 2007; Trenberth and Fasolo, 2007).
3-3-3-WRF output
Figure 11 displays MSLP and 850hPa wind data at 12 UTC on June 9 and 11, 2023. On June 9, the low-pressure system associated with TC Biparjoy is evident in the eastern AS. The wind speed around and south of the cyclone was notable, but considerably lower in other regions. By June 11, the low-pressure system intensified and shifted northeastward, positioned close to northwest India and southern Pakistan. One contributing factor to the movement of TC towards the coasts of India and Pakistan was the presence of a depression in this region, as TCs tend to move towards low-pressure centers (Zhong, 2018). Furthermore, the movement of the TC to the right and its approach towards the coasts might also be attributed to the Coriolis force. Numerous studies have demonstrated that the Coriolis force, resulting from the Earth's rotation, significantly influences the movement direction of TCs (e.g. Chan, 2010; Deng and Li, 2020). On this particular day, the low-pressure extended into the Oman Sea and the Strait of Hormuz, leading to a substantial decrease in pressure in these areas. The northerly winds along the northern shores of the Oman Sea acted as a barrier, restricting sufficient moisture penetration into the southeast of Iran.
The 700hPa RH at 00 UTC on June 11, 2023 is shown in Fig. 12-a. The center of maximum RH was situated in the east of the AS and near the west coast of India. Notably, the eye of the storm, characterized by lower humidity, is visible within the center of maximum humidity. The eye of the storm, located at the center of a TC, is a relatively calm and clear area with light winds (Shivamoggi, 2022). Surrounding the eye is the eyewall, where the strongest winds and heaviest rainfall occur (Zhu and Yu, 2019; Wang and Wu, 2004; Wang, 2012). The intense convection and strong updrafts in the eyewall contribute to cyclone development and persistence (Zhang et al., 2012; Hendricks, 2012). Whorled bands of lightning extending outward from the eyewall are known as rain bands (Shi et al., 2020; Kepert, 2018), varying in intensity and capable of producing heavy rainfall and winds (Chen et al., 2014; Chen et al., 2010). The rain bands and the significant moisture gradient between them and the cold air falling zones are well illustrated. Figure 12-b depicts the map of wind shear between the levels of 850 and 200hPa (black and red vectors, respectively) and their differences (shaded) at 12 UTC on June 11th, 2023. The maximum wind speed was around the TC, with the wind speed close to the surface much greater than the upper levels of the troposphere. The lowest wind shear is observed over the northwest of India, influencing the TC's path. Wind pattern is a critical factor affecting TC trajectories, as TCs tend to move toward areas of low wind shear, as large amounts of wind shear can disrupt or weaken their development (Zhang and Tao, 2013; Reimer and Montgomery, 2011). Above the storm center, there is an upper-level outflow, where air diverges and flows outward (Sears and Velden, 2014; Rappin et al., 2011). This outflow is crucial for maintaining the vertical structure of the storm and its development (Rohli et al., 2021). Strong wind patterns can disturb the outflow and hinder storm growth (Wong and Chan, 2004; Finocchio and Ríos Berrios, 2021)."
Figure 13-a, illustrates the 850hPa vertical flux of RH at 12 UTC on June 10th, 2023. Negative values of the flux within the eye of the storm indicate downward movements, while positive and negative values in the rain bands and falling areas can be observed, respectively. In Fig. 13-b, the CAPE on the same day is displayed. The CAPE value in the eye region was greater than in the surrounding areas. In the regions around the eyewall, the CAPE amount decreased up to a certain radius due to the high speed and configuration of the wind, significant humidity, and heavy rainfall. The greatest amount of CAPE is observed in the outer rain bands, which were away from the downdrafts around the TC.
In the exploration of dynamic instabilities in the atmosphere, various quantities, including vorticity and potential vorticity, are commonly studied (Khansalari et al., 2020; Borhani et al., 2022). Additionally, a quantity known as helicity serves as an indicator of atmospheric vortex strength, applicable to phenomena such as supercell thunderstorms, TCs, tornadoes, and dust devils (Hide, 1989 and 2002; Kurgansky, 2008 and 2017; Koprov, 2015; Farahani et al., 2017; Li et al., 2022). Among the thermodynamic and dynamic factors considered, low-level environmental helicity exhibits the strongest positive correlation with the size of TC outer-core (Li et al., 2022). Therefore, Storm Relative Environmental Helicity, a crucial dynamic factor, provides insight into the size of the TC's outer-core.
The helicity map in the three-kilometer layer above the Earth's surface at 1200 UTC on June 10, 2023 (Fig. 14a) clearly indicates the presence of a TC. In this TC, the low-level jet speed was higher compared to the upper-level jets (Figs. 14b, c, d), signifying the primary source of the cyclone's energy. The height of the eye wall of this TC extended to near the tropopause (12 kilometers above the Earth's surface). Additionally, the simulated maximum potential vorticity in this TC was approximately 36 PVU. The vertical structure of potential vorticity, potential temperature, and relative humidity depicts tropopause folding in the upper and middle levels of the troposphere. This vertical structure was examined on the center of the simulated TC in four directions: west-east (90 degrees), north-south (0 degrees), and northeast-southwest and northwest-southeast (135 and 45 degrees). In Fig. 14, the vertical structure is simulated for four directions centered on TC Biparjoy, located at a geographical latitude of 17⸰N and longitude of 68⸰E. The simulation is presented for 1200 UTC on June 10, 2023.
3-3-4- Comparison of simulated rainfall with satellite data
In Fig. 15-a, depicting the daily rainfall scatter diagram of GPM data and WRF model output from June 6 to 12, 2023, most points are situated in the bisector zone, suggesting that the model's accuracy in simulating this storm is acceptable. However, it is noteworthy that numerous points are positioned close to the horizontal axis, indicating that in these regions, the satellite data recorded low rainfall amounts, whereas the model output estimated higher values.
Figure 15-b presents the cumulative daily rainfall from GPM satellite data and WRF model output on June 6 and 12, 2023. On June 6, according to GPM data, the maximum rainfall occurred in the south of the AS. While the WRF model accurately simulated the location of the maximum rainfall, it underestimated the intensity and extent of the rainfall core. On June 12, the rain core once again moved northeastward, approaching the northwest coast of India. However, on this day, the model output misrepresented rainfall across the eastern strip of the AS, adjacent to the coast of India. For Biparjoy, the model correctly simulated the direction and speed of the storm but showed rainfall in a broader area around the storm's center.
3–4 Tropical cyclone Shaheen
Figure S2 displays the true color image of TC Shaheen from September 30 to October 3, 2021. On September 30, the Shaheen storm was positioned on the eastern coast of the AS. Gradually moving westward, on October 1st, its cloud mass entered southern Pakistan and southeastern Iran. By October 2nd, Shaheen was located in the Oman Sea, and its cloud mass covered the southeast regions of Iran and the northeast of Oman. Continuing its westward trajectory, the storm's intensity decreased, and on October 3rd, it moved from the Strait of Hormuz to the southern regions of the UAE and northern Oman. Satellite images indicate that Shaheen had a higher speed than Biparjoy, and this storm had a shorter lifetime.
Figure S3 illustrates the cloud effective radius in both water and ice phases, along with the optical depth from MODIS/TERRA on October 1, 2021. In the storm zone, ice-phase clouds with an effective radius ranging from about 30 to 35µm are observed. Additionally, water-phase clouds with a notable effective radius (between 15 and 30µm) surrounded the storm. The Cloud Optical Depth in the ice phase was higher at distances closer to the eye of the storm, with some areas reaching values exceeding 100. In the north of the storm, clouds were predominantly in the liquid phase, and their optical depth surpassed that of the clouds in the south.
The CALIPSO satellite traversed the Oman Sea at approximately 2200 UTC on October 2nd, coinciding with the path of TC Shaheen. Figure S4 presents some of the satellite's products. In the region between 25 °N and 58 °E, situated in the Oman Sea, at an altitude ranging from 7 to 15km, Total Attenuated Backscatter values exhibited a significant increase, indicating the presence of substantial clouds in this area. A decrease in potential temperature was noted in this region, possibly attributed to the release of latent heat during the cloud formation process.In an area around 20 °N in the country of Oman, tropopause folding occurred, accompanied by the descent of cold stratospheric air, resulting in the formation of high-level clouds in this vicinity. The Cloud Phase product reveals that the majority of clouds over the Oman Sea were in the ice phase and were notably thick. Below the ice-phase clouds, clouds in the water phase are observed at heights between 3 and 8 km. The cloud type product indicates that over the Oman Sea, the clouds were deep convective, penetrating up to a height of approximately 15 km in the atmosphere, with cirrus clouds evident around and above them.
The long-term MSLP pattern (Figure S5-b) reveals a band of low pressure extending below 28 °N, encompassing the AS, Indian Ocean, India, Pakistan, the Arabian Peninsula, and southern Iran. During the occurrence of TC Shaheen, there was an observed strengthening of the low-pressure system over these mentioned regions, resulting in a decrease in pressure values. Consequently, a low pressure of 1004hPa predominated over the Oman Sea, the Persian Gulf, the country of Oman, and Eastern Saudi Arabia (Figure S5-c). The highest negative pressure anomaly of -8hPa is observed over the Oman Sea, the Persian Gulf, and the country of Oman (Figure S5-a).
In the long-term 850hPa wind pattern for October, a weak LLJ is observable along the southern coasts of the Indian peninsula and in the NIO with a westerly flow. The northern current in the west of Afghanistan extended toward the Oman Sea, further manifesting as northeast currents along the coasts of Yemen, Oman, and Saudi Arabia towards the Red Sea (Figure S6-b). This pattern significantly intensified during TC Shaheen, extending from the west of Afghanistan to the eastern border of Iran and continuing towards the Oman Sea, Hormuz Strait, and Saudi Arabia. Concurrently, the LLJ strengthened in the Horn of Africa. The LLJ persisted in both directions; the left exit winds blew towards the north of the AS, and the right exit winds blew towards the southern coasts of India and the south of the Bay of Bengal (Figure S6-c). Consequently, a positive wind anomaly with a southwesterly flow is observed from the Horn of Africa to the center and north of the AS. The northerly currents from eastern Iran and western Afghanistan to the shores of the Oman Sea were impeded due to the influence of the southernly currents of the AS (Figure S6-a).
The maximum positive anomaly of the 500hPa potential vorticity is observed over the Oman Sea (Figure S7-a). During Shaheen, the SST in the Bay of Bengal and Northern Indian Ocean increased by approximately + 1°C, while in the Oman Sea and the Persian Gulf, it experienced a more significant increase ranging from + 1 to + 2°C. Conversely, there was a decrease in SST of -1 to -3°C observed in the western Arabian Sea (the Horn of Africa) (Figure S7-b).
MSLP and 850hPa wind from the WRF model output at 12 UTC on October 1 and 3, 2021, are illustrated in Figure S8. On October 1st, a low-pressure system is observed in the northeast of the AS. Around this system, high wind speeds and the evident rotation of wind vectors can be observed, while the wind speed was considerably lower in areas further away. By October 3, the low-pressure system entered the Oman Sea and weakened. One contributing factor to the transfer of the TC towards Oman was the presence of a low-pressure system over this region.
Figure S9-a displays the 700hPa RH. On October 2, above the surface cyclone, RH sharply increased, indicating significant cloudiness. The humidity gradient between the rain bands and the areas of cold air is well pronounced. Figure S9-b depicts wind vectors at 200hPa (in red), 850hPa (in black), and their differences (shaded) on October 2, 2021. Around the TC, there was relatively high wind shear, with the 850hPa wind speed greater than that at 200hPa. In the southern regions of the AS, surface winds were westerly, while at upper levels, they were easterly. Consequently, there was a substantial difference in wind speed between these two levels in this area. Wind shear was comparatively low in the west of the TC. Wind shear is a crucial factor influencing the path of TCs, as these storms tend to move towards areas with lower wind shear.
Figure S10-a illustrates the vertical flux of RH at 950hPa on October 2, 2021. TC was situated in the northwest of the AS. Positive values in the vertical moisture flux denote the presence of rain bands, while negative values indicate the descent of colder air masses. In Figure S10-b, the distribution of CAPE on October 1, 2021, at 12 UTC is presented. More CAPE values are observed particularly around the eyewall, coinciding with the intensification of the TC. Notably, the coasts of Oman exhibited higher CAPE values compared to the surrounding areas.
In terms of helicity and potential vorticity, the Shaheen TC was weaker compared to the other two TCs in this study. Figure S11 displays the helicity map in the 3-km layer from the Earth's surface for Shaheen TC. Analyzing the recorded history of this TC and simulation outputs at 12 UTC on October 1, 2021, indicates a transition from a Cyclonic Storm (CS) to a Severe Cyclonic Storm (SCS). This transition is distinctly evident in the helicity map (Figure S11). On the specified day, the height of the TC's eyewall increased to approximately 12 kilometers from the Earth's surface, coinciding with a peak potential vorticity of 8 PVU. Notably, in the Shaheen TC, both low-level and upper-level jets contributed almost equally to the overall intensity of the cyclone.
Figure S12-a presents a scatterplot comparing GPM precipitation and WRF model output from September 29 to October 4, 2021. Notably, many points are located near the two axes, indicating instances where the WRF model tended to overestimate low precipitation amounts in specific regions and underestimate higher amounts in others. The widespread dispersion of points and their distance from the bisector collectively suggest that the model's performance in simulating the Shaheen TC differs significantly from satellite data. In Figure S12-b, the daily cumulative rainfall for GPM satellite data and WRF model output on October 1 and 3, 2021, is depicted. On October 1, satellite data positioned the precipitation mass to the north of the AS, with a discernible similarity between the model output and GPM data. However, the model failed to capture rainfall over southern Pakistan. On October 3, satellite data indicated the maximum rainfall in the western part of the AS, accompanied by heavy rainfall in southern Iran and northern Oman. In contrast, the WRF model placed the precipitation slightly to the west and in the central AS. In general, while the WRF model successfully simulated the Shaheen TC, it underestimated the storm's speed and the intensity and spatial extent of precipitation.
3–5 Tropical Storm Mekunu
The true-color images captured by the MADIS sensor on the Terra satellite from May 23 to 26, 2018 are shown in Figure S13. On May 23, 2018, the formation of TC Mekunu near the Horn of Africa in the western Arabian Sea is evident, with the associated cloud mass clearly visible in the satellite image. Subsequently, on May 24, the storm shifted slightly northward, leading to an expansion of the surrounding cloud mass. By May 25, the TC continued its trajectory toward the northwest, nearing the coasts of Oman and Yemen. On May 26, the cloud mass fully enveloped over the country of Oman.
In Figure S14, the effective radius of clouds and their optical depth on May 23, 2018, is depicted. Notably, clouds in the ice phase were discernible within the tropical storm zone, with their effective radius increasing as they moved away from the storm's center. Continuing along the rain band situated to the east of TC Mekunu, clouds exhibiting water phase are evident, characterized by an effective radius ranging from approximately 15 to 30 micrometers. The highest values of COD were concentrated around the eyewall, with a noticeable decrease in COD values as the distance from the storm's eye increases.
Around 09 UTC on May 23, 2018, the CALIPSO satellite traversed the western AS, closely skirting the Horn of Africa and extending towards the country of Oman, as illustrated in Figure S15. Notably, in the region between 9 and 15⸰N latitude and at altitudes ranging from 11 to 17km, Total Attenuated Backscatter values exhibited a substantial increase. This elevation suggested the presence of clouds in the specified region. Simultaneously, there was a noticeable decrease in potential temperature, indicative of the potential influence of latent heat release during the cloud formation process. Analyzing the cloud phase product reveals that the clouds in this area predominantly existed in the ice phase. Furthermore, the majority of these clouds exhibited characteristics of deep convection, with cirrus clouds observed in proximity.
As depicted in Figure S16, the MSLP for May reveals a low-pressure system spanning the Bay of Bengal (BOB), the Indian Peninsula, Pakistan, southern Afghanistan, southeastern Iran, and the AS. Notably, a 1000hPa low-pressure center was situated in eastern Pakistan (Figure S16-b). During the occurrence of TC Mekunu, two distinct low-pressure systems emerged. The first was positioned over India and East Pakistan, with a 1000hPa center located over North India. The second was observed along the coast of the Horn of Africa, extending into the Gulf of Aden and the western AS, featuring a 1000hPa center (Figure S16-c). Examining the pressure anomaly map (Figure S16-a) reveals a noteworthy decrease in pressure in the western AS and the Horn of Africa, indicating the formation of low-pressure systems in this region. In contrast, there was an observable increase in pressure over the Himalayas, northern Afghanistan, and northern and western Iran.
The 850hPa long-term wind on May presents a weak LLJ over East Africa, which flows in the north-east direction towards the AS; anticyclone was over Saudi Arabia and northerly winds blowing from western Afghanistan and eastern Iran to southeastern Iran and continued to southern Pakistan and northern India. The second part of the jet was on the southern coasts of the India in the western direction towards the Bay of Bangla (Figure S17-b). During the Mekunu, the LLJ was strengthened over the Horn of Africa; the left exist region of LLJ was diverted towards the country of Oman with a counter-clockwise circulation, and Arabian anticyclone also was strengthened. The northerly wind blowing from west of Afghanistan and east of Iran, which was caused by high pressure over the country of Turkmenistan, was strengthened and continued to the east of the Oman Sea and then to the east of AS. This wind flow, together with the western flow of the right exit winds of LLJ, blew towards the southern coasts of the India and the Bay of Bangla (Figure S17-c). It can be seen in the anomaly pattern that there was a positive wind anomaly with an eastward flow from the AS to the country of Oman. In addition, the northerly flow from the eastern Iran and western Afghanistan was strengthened due to the high pressure over the country of Turkmenistan (Figure S17-a).
The 500hPa potential vorticity anomaly map (Figure S18-a) reveals maximum positive values over the AS in the latitude range of 10 to 18°N. During the period of TC Mekunu, SST anomalies illustrated distinct patterns. There was a noticeable temperature increase in the Bay of Bengal ranging from + 1 to + 2°C, as well as in the AS and the Persian Gulf, where anomalies were around + 1°C, and exceeding + 2°C in the northern AS. Conversely, in the southwestern AS along the Horn of Africa coast, there was a marked decrease in SST anomalies ranging from − 1 to -3°C (Figure S18-b).
In Figure S19, MSLP and 85 hPa wind data at 1200 UTC on May 23 and 26, 2018, are presented. On May 23, TC Mekunu was situated in the southwestern AS near the Horn of Africa. Notably, the wind speed around the cyclone was exceptionally high, with observable rotation in the wind vectors. By May 26, the cyclone made landfall in the southern regions of Oman, leading to a slight decrease in wind speed after the landfall event.
Figure S20-a, depicts RH at 700hPa at 1200 UTC on May 23, 2018. On this date, the maximum RH was observed in the southwestern AS. Notably, within this maximum humidity center, a distinct and sharp decrease is evident, signaling the presence of the eye of TC. In Figure S20-b, the wind pattern between 200 and 850hPa levels at 12UTC on May 23, 2018, is illustrated. The wind shear was notable, attributed to higher wind speeds at 850hPa compared to the upper levels of the troposphere in the TC area situated near the Horn of Africa.
The vertical flux of 950hPa RH at 1200 UTC on May 23, 2018 depicted in Figure S21-a, reveals both positive and negative values around the eye of the TC. Positive values correspond to the presence of rain bands, while negative values indicate areas of cold air descent in this region. Figure S21-b displays the CAPE values on the same day. Notably, the CAPE values were lower in the proximity of the storm's eye and higher in the rain bands, particularly those situated farther away from the center of the storm.
As depicted in Figure S22, during the peak activity of TC Mekunu on May 23, 2018, helicity values reached their maximum. Subsequently, by the late hours of May 25, the TC made landfall on the shores of Oman, gradually diminishing in intensity (not illustrated in the figure). An examination of the vertical structure of potential vorticity, RH, and potential temperature unmistakably reveals the presence of the TC's eye wall. Within the eye wall region, RH exceeded 80%, extending to an altitude exceeding 10 kilometers. Additionally, potential vorticity displays two maximum values—one near the Earth's surface and the other in the upper levels of the troposphere. This dual-peak pattern is a consequence of the increased speed of the low-level jet relative to the upper-level jets during the peak days of TC activity, signifying the primary source of energy for the TC—the release of latent heat from the warm waters of the AS and the Indian Ocean. The structural depiction of potential temperature, potential vorticity, and relative humidity illustrates tropopause folding in the upper and middle levels of the troposphere. At 12 UTC on May 23, when the TC transitioned from a Cyclonic Storm (CS) to a Very Severe Cyclonic Storm (VSCS), the potential vorticity value near the Earth's surface (approximately 2 kilometers above the Earth's surface) reached an exceptionally high value of about 46 PVU.
In Figure S23-a, the scatter plot depicts the daily cumulative precipitation of the GPM satellite product and WRF model output from May 20 to May 26, 2018. Most points on the graph closely align with the bisector, indicating an acceptable accuracy of the model in estimating precipitation. However, there is a slightly higher number of points located above the bisector than below, suggesting instances where the model underestimated precipitation.Figure S23-b illustrates the daily precipitation of satellite data and WRF model output on May 21 and May 25, 2018, pertaining to TC Mekunu. On May 21, the satellite data reveals a substantial rainfall mass near the Horn of Africa, consistent with the WRF output. However, the WRF model shows rainfall over a larger area in the southeast of the AS not present in the satellite data. On May 25, the rain associated with Mekunu made landfall in the south of Oman and along the border with Yemen, with a reduction in the extent of rainfall around the storm. The WRF model accurately simulated the landfall and intensity of precipitation, but it overestimated the area with less rainfall.