1 Introduction

The troposphere is the atmosphere's lowest layer. Weather, cyclones, and anticyclones are all part of it. It stretches from the earth's surface to a height of roughly 15 km, with the upper limit fluctuating depending on the site's geographical position (i.e., from pole to equator) (Stull 1988). Its permanent constituents are nitrogen, oxygen, argon, and carbon dioxide, with their relative concentrations essentially unaltered (Kumar et al. 2021a). In addition, the troposphere contains several additional changeable elements, the most prominent of which are ozone and water vapour. Near the earth's surface, the largest concentration of water vapour can also be observed. Wind speed is also shown to rise with height, but pressure falls consistently with height, temperature decreases less consistently, and humidity decreases quite erratically (Stull 1988). The atmospheric boundary layer (ABL) is the lowest portion of the troposphere where the direct effect of surface heating and cooling can be seen (Kumar et al. 2021a). All biological and human activities take place within this layer. The ABL extends to the height of convective mixing of buoyant parcels of air or the base of an elevated inversion layer under conditions of strong upward heat flux, i.e., under daytime unstable conditions, and the level of surface-based inversions determines the height of the ABL during night-time, i.e., under stable conditions (Kumar et al. 2017a). Turbulence is one of the ABL's most important components. It allows transporting water vapour, heat, momentum, contaminants, and other tracer components from one location to another.

Many factors control the air quality at a place at any time of the day (Garratt 1994). Some of these factors are constant for a particular location, while others are variable. The constant ones include the topography and climatology of the place, nature of the pollution sources in and around that place, nature and amount of the effluent and the demography of that location (Bontempi and Coccia 2021). The variable factors are the time distribution of emission, synoptic weather pattern, meso-scale low-level stability, inversion topography and depth, and flow pattern. Careful examination of meteorological parameters reveals that only two factors can be classified as key in managing pollution concentrations in a specific area with a set of constant controls. These are the low-level temperature structure and low-level wind direction and strength (Stull 1988). The low-level temperature structure determines whether the atmosphere is in a state of static stability or instability and under these conditions, diffusion of pollutants is either suppressed or enhanced (Liao et al. 2021). The wind velocity, on the other hand, determines the path the pollutants will take after the emission and how rapidly the effluents will be carried away. These conditions are thus a necessary input to determine whether a given emission rate will produce a critical concentration level over a region.

In light of the importance of the atmospheric structure and dynamic processes in the lower atmosphere to a wide range of human endeavours, various types of monitoring techniques have been developed over time to monitor the atmospheric parameters such as temperature, humidity, wind speed and direction, and their distribution in space (Stull 1988). In-situ procedures and remote sensing techniques are the two primary categories for these methods. The details of these techniques are mentioned by Seibert et al. (2000).

SODAR (SOnic Detection And Ranging) is a device based on remote sensing to monitor the ABL and their structure (Kumar et al. 2014, 2021b, c). SODAR has the best refractive index among other remote sensing instruments. It is internationally recognised and recommended by the Environmental Protection Agency (EPA) for air quality modelling in the Environmental Impact (EIA) Assessment (Stull 1988, EPA Report 2000). SODAR is the cost-effective instrument for real-time monitoring of ABL heights up to 1–2 km. The Bhopal Gas Tragedy attracts the scientific community to study ABL to measure the VC of an area. Given the conditions, the Central Pollution Control Board (CPCB) Delhi has been included SODAR system in the list of mandatory equipment for air quality management to avoid such disasters (CPCB Report 1992).

SODAR is one of the best remote sensing systems recognised worldwide and has proven effective in providing real-time data on air pollution controls such as mixing height (ABL height)/altitude conversion. This is the first time that the SODAR system has been installed in Mumbai region (near the coastal region, Sion). It is the first time that such an analysis using ABL height real-time hourly data is being performed over Mumbai region. SODAR provides useful information for the investigation of parameters relevant to air pollution studies. In the present paper, different structures of ABL height along with ventilation coefficient have been studied, and ABL height comparison with meteorological parameters also discussed.

2 Geographical study of the Mumbai region and monostatic SODAR details

Increasing air pollution in Mumbai city is one of the significant problems for policymakers. The geography of Mumbai results in the advection of air into the city from the surrounding areas, which can sometimes be significantly more polluted than the city centre itself (Kumar et al. 2016). Mumbai is on a narrow peninsula on the southwest of Salsette Island, which lies between the Arabian Sea to the west, Thane Creek to the east and Vasai Creek to the north (figure 1).

Figure 1
figure 1

Map of Mumbai City (courtesy: www.mapindia.com).

Full size image

2.1 Details of MPCB SODAR system

MPCB has deployed a monostatic SODAR at Mumbai that provides information on the thermal structure of the lower atmosphere. The installation of the SODAR at the MPCB headquarters is completed in November 2020. The study location is at the intersection of 72.865°E longitude and 19.042°N latitude. As per the topography, the area is located far away from the sea is ~3 km. The atmospheric turbulent eddy diameters that cause acoustic wave scattering are on the order of half a wavelength. The SODAR provides information about thermal plumes, inversion layer breaking periods, boundary layer heights, mixing layers, and other related phenomena in time and space.

Based on the foregoing considerations, the backscattering monostatic SODAR system is designed and developed (Aggarwal et al. 1980; Singal et al. 1985; Gera et al. 2011; Kumar et al. 2021a, b). Figure 2 shows a schematic block diagram for the monostatic SODAR. It is used in a monostatic mode antenna emitting high power acoustic burst of 100 ms duration every six seconds at acoustic centred frequency 2250 Hz (Kallistratova 1963; Brown and Hall 1978). The technique is used by transmitting a pulsed narrow sound beam into the atmosphere, where air inhomogeneity is contacted and partially reflected. It has received dispensed signals from the same transducer. With each echo scan, the time of delay and intensity are measured as intensity module on the PC, with visually displayed in either operational mode the height range (ordinate) vs. the time (abscissa). The two main kinds of acoustic echoes seen in echograms are inversion echoes and thermal echoes. Thermal echoes appeared as vertical intermitted spikes, whereas inversion echoes were horizontal and continuous in time. An acoustic wave propagating through a turbulent medium gets scattered, refracted, and attenuated. The basic data is acquired by using the monostatic SODAR system. The operation parameters are shown in table 1.

Figure 2
figure 2

Block diagram of SODAR system.

Full size image
Table 1 Specifications of monostatic SODAR system.
Full size table

In the new SODAR system, an Ultra-Low-Noise Amplifier (ULNA) and Dual-Amplifier Band Pass (DABP) filter are designed for analog signal conditioning of SODAR (Chourey et al. 2022a, b). The updated analog signal conditioning system features improved gain and high signal-to-noise ratio (SNR) of 65 and 92 dB, respectively. The results were then compared with the existing analog signal conditioning system. It was observed that the updated system has significant advantages in terms of gain and narrow bandpass response. Additionally, the updated SODAR system captured the echogram structures. The results demonstrate the efficiency of the echogram in capturing the well-known standard structures of ABL.

A unique software is developed for monitoring and storing all the data observed by SODAR on a LABVIEW platform (online software and offline software) (Kumar et al. 2021b, c). The system consists of two software. (1) Online Software, which is responsible for running the SODAR system, acquiring and logging FAX plots as well as weather data. (2) Offline Software, wherein the user can view historical SODAR data, do detailed studies as well generate reports.

This new updated version of the SODAR system has a set of meteorological sensors integrated into it, thus providing valuable insight into temperature, humidity, wind speed and wind direction data, which can be related to the ABL height variations in the lower atmosphere. The system is capable of monitoring the ABL up to a height of 1000 m. The observation period from 5th November 2020 to 31st October 2021 is used in this study.

2.2 ABL height determination from SODAR echogram

The SODAR is a cost-effective remote sensing system that uses sound waves. It allows 24×7 monitoring. A SODAR emits sound pulses at various heights of the atmosphere and receives back dispersed pulses in temperature inhomogeneities. SODAR echograms provide lower climate turbulence pictures as well as contaminant distribution responsibilities. Based on the vertical profile of the acoustic refractive index shown in figure 3, the echogram (9th May 2021), atmospheric conditions are divided into two categories: convective period (unstable) and non-convective (stable) period (output of SODAR). Beyrich (1997), Singal et al. (1997) and Bradley (2007) have compiled methods and algorithms for calculating the ABL height using SODAR data. Being based on sonic principles, the range of SODARs is limited to a few hundred meters, extending a maximum of up to 1000 m based on the power and frequency of the emitted pulse. The observer needs to broadly classify the data into two categories, namely: inversion-time and convection-time. The other classifications of ABL structure are fog-layer, rising-layer, multi-layer, etc., and each classified structure demands a different approach for the ABL height estimation. The structures are manifestations of different prevailing meteorological conditions. The meteorological processes responsible for generating such structural SODAR signatures include phenomena such as free/forced convection, nocturnal cooling of ground, sea breeze, land breeze, advection, subsidence, frontal system, etc. These phenomena generate a turbulence zone boundary in ABL, which in turn acts as a tracer for SODAR interrogation.

Figure 3
figure 3

Stable and convective period on SODAR Echogram (9th May 2021).

Full size image

The ABL height has been directly picked up from the echogram by using visualisation, apart from the convection period. A measurement of the height of the thermal plumes by SODAR during the day will always give an underestimated value unless they are capped by a low-level elevated shear echo layer; however, as SODAR sensitivity declines in the daytime compared to the night-time due to the prevalent ambient noise which decreases its probing range. Based on a Holzworth model and radiosonde data for Delhi, a method to estimate mixing height during the daytime when the plumes are not covered by a stable layer has been developed (Singal 1993; Singal et al. 1994). Therefore, the actual ABL height is determined by using an empirical formula equation (1) (Singal et al. 1985).

$$y=4.24x+95,$$
(1)

where y is the calculated ABL height and x is the observed thermal plume height in the echogram. This formula was determined by comparing SODAR data with Radiosonde data (Singal and Aggarwal 1979).

3 Results and discussion

3.1 Sea breeze and their signature on SODAR echogram

Sea breeze is a phenomenon occurring in the coastal region and is associated with the manifestation of onshore local winds which set in from the sea onto land in the late forenoon or early afternoon. The phenomenon is marked by drop in surface temperature, rise in humidity level, change in wind direction, increase in wind speed and suppression of convective activity. Temperature differences between land and water surfaces give rise to the circulation of land and sea breeze. Sea breeze is especially well developed on a clear sunny day under conditions of calm to light winds. The land–sea temperature contrast is usually reversed at night, resulting in land breeze, which is weaker than the daytime sea breeze.

The turbulent changes in the ABL due to the development of sea breeze during the daytime and land breeze during night-time give their characteristics signatures on the SODAR echograms. Sea breeze is detected on the SODAR echograms many times through the formation of capping layer above the daytime slightly diffused thermal plume structure, soon after the erosion of the morning rising inversion layer. Aggarwal et al. (1980) studied the atmospheric structure using SODAR and meteorological tower in 1978 at Tarapur and found that thermal convective structures are seen during the daytime and at night time. The height of convective boundary layer in the daytime is 400–500 m, while at night-time, it is mostly under 200 m. Singal (1991) also studied the sea breeze circulation on the Indian coastline (Tarapur and Visakhapatnam) during 1990–1991 and found that continuous real-time data of onset time, duration, and depth of the internal boundary layer during the flow of sea breeze at a coastal site. Gera et al. (2013) studied the coastal boundary layer characteristics using the SODAR in the coastal region of the Arabian Sea, Antarctica Ocean and Indian Ocean in 1999, 2001 and 2007, respectively. They found that anomalous thin elevated layers oscillating at periods of about 24 hours in the coastal region.

The phenomena of the sea breeze complexity sparked a lot of interest in understanding the lower troposphere dynamics during sea-breeze events. In coastal areas, the system's sea breeze behaviour and inherent stratification, particularly the Thermal Internal Boundary Layer (TIBL), are critical meteorological elements (Talbot et al. 2007). Furthermore, knowing the sea wind is necessary for air quality forecasts and pollution monitoring in highly industrialised and densely populated metropolitan coastal locations (Beyrich et al. 1996). During sea breeze situations, TIBL is formed over land with different diffusive characteristics inside and above the TIBL layer (Prabha et al. 2002). Prabha et al. (2002) also studied the thermal and dynamical influence on the TIBL, which is altered during different synoptic winds, i.e., changing synoptic conditions from winter to summer pressure patterns over the Indian subcontinent. The onset time and duration of sea breeze are closely associated with the synoptic forcing (Kirankumar et al. 2019). Updrafts associated with the sea breeze front are found to be stronger, and turbulence intensity is lower during the weak onshore synoptic wind case (Iwai et al. 2011). Reddy et al. (2021) studied the characteristics of the sea breeze and TIBL during different seasons at Kattankulathur and Meenambakkam over Chennai and observed that TIBL height is found to be at altitude ~0.54 ± 0.16 and 0.62 ± 0.13 km at Meenambakkam and Kattankulathur, respectively. TIBL shows a weak seasonal variation with maximum height during winter and minimum height during summer in contrast to the seasonal variation of the CBL.

In the present study, the ABL height has been estimated using the SODAR echogram as a function of time. Particularly in the months of November, December, January, February, August, September, and October, the ABL height often approached 700 m. The ABL height is increased to 1000–1200 m in April because of the strong wind and temperature. Except for April month, the atmosphere is generally characterised by thermal convection from 0900 to 1800 and thermal inversion from 1900 to 0800. Since there were a large number of ecograms observed throughout the study period, certain ecograms from various seasons are included in this study as it is not possible to represent all ecograms. Figures 4–8 show the structural changes in ABL for the dates of November 5th, 2020 (clear day), November 16th, 2020 (cloudy day), December 8th, 2020 (clear day), April 5th, 2021 (clear day), and September 20th 2021 (cloudy day), respectively. Tables 2–6, on the other hand, include numerical values for ABL height, relative humidity, temperature, wind speed, and wind directions for the same days, respectively.

Figure 4
figure 4

SODAR Echogram (5th November 2020).

Full size image
Figure 5
figure 5

SODAR Echogram (16th November 2020).

Full size image
Figure 6
figure 6

SODAR Echogram (08/12/2020).

Full size image
Figure 7
figure 7

SODAR Echogram (05/04/2021).

Full size image
Figure 8
figure 8

SODAR Echogram (20/09/2021).

Full size image
Table 2 SODAR data (ABL height with meteorological data) 05/11/2020.
Full size table
Table 3 SODAR data (ABL height with meteorological data) 16/11/2020.
Full size table
Table 4 SODAR data (ABL height with meteorological data) 08/12/2020.
Full size table
Table 5 SODAR data (ABL height with meteorological data) 05/04/2021.
Full size table
Table 6 SODAR data (ABL height with meteorological data) 20/09/2021.
Full size table

3.2 Temporal variability of ABL height, ventilation coefficient, and wind speed

Pollutant dispersal capacity and the structure of the lower atmosphere are determined by the ABL height. The higher the ABL height, the larger the dispersion rate. SODAR echograms show that the ABL height is constantly changing. As a result, the information is conveyed using a box plot. Data values are displayed using a box plot, which shows the level, dispersion, and symmetry of values by plotting data points along the median, quartiles, and lowest and highest points. Every box contains a central mark, which represents the median value, while the bottom line represents the 25th percentile and the top line indicates the 75th percentile of the data. Each outlier is represented by ‘+’ symbol, and the most significant data points are covered by whiskers. An example of temporal ABL height variability is shown in figure 8. The ‘±σ’ standard deviation is shown by the vertical bars. Figure 9 and table 7 show the temporal and monthly ABL height average SODAR data over a period from November 2020 to October 2021. Higher convection period is observed during April (about 1290 m) and the lowest during February (about 330 m). The monthly average maximum value (418 m) of height is measured in April, whereas the lowest value of ABL height (179 m) is found in February month.

Figure 9
figure 9figure 9

Monthly temporal plot of ABL height.

Full size image
Table 7 Monthly ABL height variation during different hours.
Full size table

Figure 10 and table 8 show the temporal and monthly average wind data over a period of nearly a year (November 2020 to October 2021). It is observed that the convection period is most prolonged in April month with a wind speed 7.2 m/s and lowest during December month with a wind speed 1.75 m/s. The monthly average maximum value (4.91 m/s) of wind speed has been observed in July, while the minimum value of wind speed (1.07 m/s) is found during December. Figure 11 shows the monthly wind rose diagram of the above period. It is observed that mostly wind flows in northwest or west direction, but in July month, it flows in only the west direction. The calm wind was absent or less from April to July.

Figure 10
figure 10figure 10

Monthly temporal plot of wind speed.

Full size image
Table 8 Monthly wind speed variation during different hours.
Full size table
Figure 11
figure 11figure 11

Monthly wind rose diagram.

Full size image

The VC is an atmospheric dispersion value which gives an indication of the air quality and pollution potential, i.e., the capability of the atmosphere to dilute and disperse the pollutants over a location (Sujatha et al. 2016). It is calculated as the product of ABL height and the average wind speed. For better atmosphere quality, the value of ventilation coefficient remains high, i.e., the more efficiently the atmosphere can disperse the pollutants (Kumar et al. 2017b). Lower VC values, on the other hand, result in poor pollutant dispersal, stagnation, and poor air quality, potentially resulting in pollution-related problems.

$$\mathrm{VC}=\mathrm{ABL \; height }\times \mathrm{average \; wind \; speed}.$$
(2)

The VC is proportional to the height of the ABL and the average wind speed. A change in the VC is caused by variations in the ABL height and average wind speed (Murthy et al. 2020). VC less than 6000 m2/s during the afternoon indicates high pollution potential, whereas ABL height of less than 500 m during the morning indicates high pollution potential (Holzworth 1972; Saha et al. 2019).

Figure 12 and table 9 show the temporal variation of VC from November 2020 to October 2021. It has been found that overall VC was higher during April (6000–7000 m2/s) due to high ABL height (figure 9) and more change in wind direction (figure 11), and lower during December. During November, December, January, February, and September months VC values peak varied from 2000 to 3000 m2/s (convection period); however, during March, May, July, August, and October, 4000–5000 m2/s VC values are recorded in convection period. Similarly, in September, VC values between 3000 and 4000 m2/s are observed during the convection period. Consequently, severe pollution loading conditions occurred from November to February. Sujatha et al. (2016) stated that the dispersion of pollutants is dependent upon many meteorological parameters, most significant being wind and ABL height that defines the volume of air through which the pollutant is mixed. As a result, the ventilation coefficient is higher during the daytime and lower at night-time, indicating that the pollution loading capacity of the atmosphere is good during the daytime and poor at night-time.

Figure 12
figure 12figure 12

Monthly temporal plot of ventilation coefficient.

Full size image
Table 9 Monthly ventilation coefficient variation during different hours.
Full size table

According to the above results, the ventilation coefficient between daytime and night-time has been significantly different. The ventilation coefficient has been shown in figure 12 and table 9 to tend to be greater and considerably variable during the daytime, while it was comparatively lower and steady during the night. This might be a consequence of the combined impact of ABL height and wind speed. The existence of ground-based inversions, which impeded the dispersion, may account for the high ABL height values that occur in the afternoon (12:00–14:00), which are mostly attributable to the maximum during the late night and early morning hours (Stull 1988). During pre-monsoon months (April and May) in the afternoon period, high wind speed was observed in comparison to post-monsoon and winter months (September–January). Similar results have also been found for average ABL height, which has been highest during the pre-monsoon months (April and May). These findings showed that daytime ventilation coefficients have been considerably greater than night-time ones, which is favourable for the dispersion of air pollution. While at night, the ventilation coefficient has been seen to be low, which may impede the dispersion of air contaminants and thus lead to poor air quality. It is observed from figures 8 and 9 that afternoons (12:00–14:00 hrs) have very high levels of both vertical (ABL height) and horizontal (wind speed) ventilation. Nocturnal wind speeds in this study varied from roughly 1–2 m/s, while afternoon wind speeds ranged from 4 to 7.2 m/s, indicating that wind speed in Mumbai is higher during the daytime than at night-time. Krishna et al. (2004) also investigated the assimilative capacity of the Visakhapatnam basin region in the pre-monsoon and winter of 2002–2003 using the ventilation coefficient and observed that the ventilation coefficient has the lowest value in the early morning and at night, and has the greatest value in the afternoon.

3.3 Daily average variability of ABL height, ventilation coefficient, meteorological parameter, and pollutant concentrations

The ABL structure plays an important role in the formation and evolution of air pollution. The direct measurements of air-pollutant concentrations and SODAR observations have shown that the local contents of pollutants relate to the wind speed and ABL height to a greater extent than with the type of stratification. Wagner and Schafer (2017) proposed that the influence of ABL height on pollutants can be analysed by grouping ABL height data into periods and correlating classified ABL with concentrations of pollutants. To further understand the relation between the ABL height, meteorological parameter, concentration of pollutants and ventilation coefficient, the 24-hourly average value of the concentration of pollutants (PM2.5 and PM10), temperature, relative humidity, wind speed, ABL, and ventilation coefficient are presented in figures 13–19. It has been found that ABL, wind speed, temperature, relative humidity, and ventilation coefficient have anti-correlation with the concentrations of pollutants (PM2.5 and PM10). Concentrations of pollutants have increased from November 2020 to February 2021 and reached maximum value during January. Variation of ABL has an opposite correlation with the concentrations of pollutants which decrease gradually from March to September. After that, the ABL height and wind speed changed and decreased, and the concentration of pollutants increased from October 2021. The opposite direction indicates that the ABL determine the height of dispersion of the pollutants. Compared to figure 12, it is found that the pollutants have dispersed in the atmosphere during the daytime (08:00–16:00), whereas they accumulated during mid-night-time and early morning time (00:00–08:00).

Figure 13
figure 13

Daily average of PM10.

Full size image
Figure 14
figure 14

Daily average of PM2.5.

Full size image
Figure 15
figure 15

Daily average of temperature.

Full size image
Figure 16
figure 16

Daily average of relative humidity.

Full size image
Figure 17
figure 17

Daily average of wind speed.

Full size image
Figure 18
figure 18

Daily average of ABL height.

Full size image
Figure 19
figure 19

Daily average of ventilation coefficient.

Full size image

It is observed from figures 11, 13, 14, 17 and 18, that the southeast winds in Mumbai during November and December 2020 are dominant and as a result, the ABL height is lowest and concentrations of pollutants are greater. However, during January month, the dominance of northwest wind increased, dominance of southeast decreased and ABL height increased. When the northwest winds are dominant and east/southeast winds are absent or negligible during the March to June 2021 months, then ABL height is increased and the concentration of pollutants is decreased. Again, in September and October months, as southeast winds increased, ABL height decreased and concentration of pollutants also increased. Furthermore, after a sharp drop in temperature (figure 15) throughout these months, a stable and shallow layer developed and persisted at night as a result of substantial surface radiative cooling. Air pollution dispersion is suppressed by stable stratification, calm winds, and humid air conditions (figure 16) in the ABL, which encourages the generation of secondary aerosols. The decrease in northwest winds and increase in southeast winds caused pollutant accumulation in the Mumbai region.

Sengupta (2008) and Nasir and Brahmaiah (2015) have observed that the concentration of the pollutants is not constant throughout the year; its variations indicate that these pollutants are not only affected by the local source, but also remote sources in other locations. From the above result, it has been found that the concentration of pollutants has changed continuously, because of local sources as well as remote sources and meteorological conditions. Simmons et al. (1971) and Beran and Hall (1974) found that the concentration of pollutants in any area grows as the square root of the distance along the trajectory when the emission intensity is uniform. Also, the concentration rises parabolically with distance over the source area, and it falls off rapidly at first, then slowly as the air moves further away from the source. Upwind mixing and lateral diffusion can only reduce, not remove, pollutant levels no matter how far downwind they are. Solar heating and winds, to an extent, help to remove pollution, but by far, the most effective process to remove pollutants from the atmospheric air is falling rain. Rain washes away the particles and gases found in cloud drops and contaminated air. Vittal Murty et al. (1980) studied the mixing height and ventilation coefficients for urban centres in India using radiosonde and observed that April recorded the highest values of both maximum and minimum mixing heights for inland and coastal stations. Also, it has been said that higher afternoon ventilation coefficients are observed in central India and South India in April and August, respectively.

3.4 Comparison of SODAR and radiosonde measurements

Radiosonde input numerical simulations and assimilation systems are typically used to generate reanalysis datasets. These schemes ensure that the distribution of grid elements is physically and dynamically reasonable. Each reanalysis dataset relies heavily on empirical parameterisations and the diagnostic method chosen, which leads to varying distributions. ERA5 is the fifth generation of ECMWF's (European Centre for Medium-Range Weather Forecasts) global weather and climate reanalysis product. Compared with the previous version ERA-Interim, this dataset has improved physical drivers and data assimilation methods (Hersbach et al. 2020). The greatest advantage of ERA5 is its use of both satellite as well as traditional data for collaborative computation, inter-calibration, and processing. It has improved the quality of not only historical observations but also the quality of temporal as well as geographic coverage. ERA5 data has spatiotemporal resolution of 0.25°×0.25° and 1 h. In this manuscript, the hourly atmospheric boundary layer for a year from November 2020 to October 2021 was extracted over Mumbai.

A comparison of the monthly variation of atmospheric boundary layer height by a SODAR and ERA5 reanalysis data is presented in figure 20 and table 10. It is observed from the figure that the general pattern of variation is similar using both datasets, but the ERA5 reanalysis data values are higher than the SODAR during all months except October. Each month, diurnal variation of both ERA5 reanalysis data and SODAR dataset is given in the Supplementary file. It is found from both comparisons that Radiosonde ERA5 unstable structure height is higher during November, December, February, and March months, while January, May, June, and July months' height are almost equal. Moreover, during April, SODAR unstable structure height was higher than the ERA5 reanalysis data. On the other hand, stable structure is within 200 m during all the months in SODAR data, however, it varies between 100 and 400 m using ERA5 reanalysis data. During the night-time, surface inversions are caused by the cooling of air upon contact with the cooler ground. Therefore, the SODAR data reveals a deep surface-based layer during the late evening hours that may persist throughout the night unless disturbed by turbulence. This type of structure implies the presence of a surface-based radiation inversion layer in the lower atmosphere under stable conditions, as depicted in figures containing both SODAR record and ERA5 reanalysis data. The difference can be ascribed to turbulence, which creates a micro-meteorological structure in the lower atmosphere that is unique to a particular location. These observations suggest that turbulence is predominantly of local origin and that more turbulence is generated at and around SODAR sites (Singal and Aggarwal 1979).

Figure 20
figure 20

Monthly average variation of ERA5 reanalysis data and SODAR ABL data.

Full size image
Table 10 Monthly average, max, min values of ERA5 reanalysis data and SODAR data.
Full size table

4 Conclusion

In the present paper, ABL height variability over Mumbai's coastline area is described using SODAR-based measurements made between November 2020 and October 2021. Due to the existence of the land/sea breeze circulation, the assessment of ABL height over the coastal station is rather complicated. The key findings of the present study are summarised below:

  • The ABL height often reached 700 m, especially in November, December, January, February, August, September, and October. Due to the high wind and warmth in April, the ABL height is extended to 1000–1200 m.

  • April has the highest convection period (1290 m), while February has the lowest convection time (about 330 m). The highest monthly average height (418 m) is recorded in the month of April, whilst the lowest monthly average height (179 m) is recorded in the month of February.

  • It has been noted that the convection period is the longest in the month of April, when the wind speed is 7.2 m/s, and the shortest in the month of December, when the wind speed is 1.75 m/s.

  • The monthly average maximum wind speed (4.91 m/s) is recorded in July, while the monthly average lowest wind speed (1.07 m/s) is recorded in December.

  • The wind often blows in a northwestern or western direction, however in July, it generally blows in a westward direction. From April through July, the calm breeze is missing or seldom present.

  • The highest VC values, around 7095 m2/s are recorded in the month of April, while the lowest values around 2912 m2/s are recorded in the months of November and December.

  • It has been observed that the concentrations of pollutants have an anti-correlation with ABL, wind speed, temperature, relative humidity, and ventilation coefficient.

  • SODAR dataset is also compared with ERA5 reanalysis dataset, and there is a difference found between the dataset due to the turbulence, which creates a micro-meteorological structure in the lower atmosphere that is unique to a particular location.