Authors: Md Masudur Rahman
Categories: Review Paper, Air quality, Atmospheric pollutant gases, Epidemiology, Measurement methods, Policy makings, Satellite instruments
Source: Arabian Journal of Geosciences
Measurement of atmospheric air pollutants (aerosols, NO2, O3, CO, HCHO, and SO2) is essential for characterizing the environmental and biogeochemical process to monitor the air quality. The concerned measurement methods and instrumentations are complex due to the weak spectral characteristics and very low concentration in volume mixtures of them in the atmosphere. This narrative review focuses on the fundamental and advanced concepts of almost all the in situ and satellite-based methods and instrumentations for measuring the gaseous and other air pollutants in the earth-atmosphere interaction. This paper discusses the literature of past and present developments in the measurement methods and instrumentations by highlighting their positive and negative feedbacks. The developing history (1970–2020) of space-borne instrumentations is indicated along with their techniques and capability to compute the concentration of the atmospheric pollutants for monitoring the air quality (air pollution) and climate in large (regional to global) scale. Several applications of the satellite instruments are described in terms of some important pollutant gases, most of which are known as ambient air pollutant. This study makes some recommendations to the readers so that they can utilize the measurement methods and instrumentations to estimate the important air pollutants such as NO2, SO3, O3, CO, and aerosols as well as for setting up future guidelines and air pollution control policies regarding the epidemiology and public health issues.
The online version contains supplementary material available at 10.1007/s12517-023-11410-4.
Keywords: Atmospheric pollutant gases, Measurement methods, Satellite instruments, Air quality, Epidemiology, Policy makings
Atmospheric pollutants such as NO2, O3, CO2, CO, HCHO, SO2, and aerosols (particulate matter, PM) are mostly responsible for air pollution and highly utilized for the environmental analysis in the list of trace gases (Loubet et al. 2020). To perform any analysis within the biogeochemical and environmental processes, the measurement of the pollutant gases is essential. Basically, atmospheric pollutant concentrations and fluxes can be defined as the number of moles/molecules or joules/watts of energy or micro/milli grams of molecular mass passing through a unit area/volume per unit time. In case of molecular concentrations, the common unit of measurement is parts per million (ppm) or parts per billion (ppb) of atmospheric trace gases. For the measurement of fluxes of trace gases, the used units are nano/micro moles per meter square per second (nmolm^−2^ s^−1^ or µmolm^−2^ s^−1^), or milli/micro grams per meter square per second/day (µgm^−2^ s^−1^ or µgm^−2^d^−1^), where these units are inter-exchangeable among them based on the adopting methods (Terraglio et al. 1958; Colahan 2004; NSW 2021). So, the efficient measuring method and respective instrument with suitable unit is very important for the specific species determination in the atmosphere. In the planetary boundary layer, the fluxes of atmospheric trace gases and energy are very important because they are an integral component of the conservation of mass, i.e., the time rate of change of the atmospheric burden of different trace gases (Reifsnyder et al. 1991). Overall, the trace gas fluxes are the important metric of an ecosystem performance, which measures the mass balance of vegetation and soils, and the way of physiological responding to the environmental forces and stresses. The measurement of pollutant gases in flux based indices is complex but better results are attained in flux based calculations (Musselman et al. 2006). There is no single or simple sensor for measuring the fluxes of trace gases at desired time and space like the other atmospheric sensors that are readily available for measuring the state variable like temperature, humidity, and wind velocity (Baldocchi 2014; Rahman et al. 2019). The complexity of measurement is arisen due to the compound structure, weak spectral signature, very low concentration in volume mixtures, multi-scaled, hierarchal, sum of many non-linear feedback controls, and modulations of trace gases. So, there is no perfect methods and instrumentations for measuring the pollutant gases in the earth-atmosphere interaction but there are several potential scientific contributions in the method and instrumentation development in the world over last 60 years (Woodwell and Whittaker 1968). This paper is presenting the in situ and satellite-based measurement methods and instrumentation of air pollutant concentrations and exchange fluxes at the earth-atmosphere interaction.
The earth system is naturally complex for its formation thorough the interconnection of different components (land, atmosphere, oceans, rivers, trees, animals, and human activities) which results the measurement methods and instrumentations are facing uncertainties in estimating the atmospheric trace gases in real-time applications. The interaction between the earth and atmosphere is the prime interface that operating on the wide range of spatial–temporal scales. The earth-atmosphere interaction is critically important for driving the earth system via mass, energy, momentum fluxes, and the biogeochemical processes, where the biogeochemical cycles are considerably constrained by the climate variability and atmospheric processes (Suni et al. 2015). An iconic diagram illustrated the earth-atmosphere interaction and their basic components as shown in Fig. 1. A number of direct and indirect methods have been utilized for measuring the atmospheric trace gases for different purposes, where the main purposes are- to evaluate the energy fluxes for understanding the local and regional climate conditions; to compute the low weighted molecules for observing the environmental situations such as the atmospheric pollutions; to detect the chemical parameters in lab experiments for other purposes.
Fig. 1 The iconic diagram of the earth-atmosphere interaction
During the last 60 years, several measurement techniques have been reported with their basic principles, advantages and limitations, where in the late 1970s and early 1980s, the eddy covariance (EC) method have been developed and successfully grown in the early 1990s (Wesely et al. 1977; Anderson et al. 1984; Baldocchi 2014). In the revolution period of personal computer (PC) with high volume data storage capability, the EC method with 3-D sonic anemometer and infrared/Laser spectroscopy with solid state devices have been applied widely for measuring the earth-atmospheric interacted parameters of water-energy cycle, carbon cycle, and nitrogen cycle. In case of water-energy cycle, there are several uncertainties happened in the measurement of energy fluxes due to the unsolved issue of surface energy balance closure (EBC) (Rahman et al. 2019). The ultimate goal of this research community is to make possible and convenient measurement of the fluxes of the atmospheric trace gases and energy in everywhere and for all of the time but it is highly challenging due to the highly heterogeneous characteristics of the earth-atmosphere interaction. So, the researchers are trying to combine all the possible measurement techniques, and their networks as well as the remote sensing (satellite and aircrafts)-based techniques by parameterizing and validating the satellite data products (Heinsch et al. 2004; Yuan et al. 2007; Xiao et al. 2011; Rahman and Zhang 2019a; Luo 2021).
With the recent advancement of electronics, electromechanical and communication engineering, the instrumentations have become more potential and faithful for measuring the atmospheric pollutant gases but till there are research gaps in this field due to the complex nature of the earth-atmosphere interaction and other phenomena. This development has been occurred very potentially in the last three decades, where the development in the atmospheric physico-chemistry also been remarkable for epidemiological studies and policy making on public health. So far, we know that there are rare review articles during the last decade in this specific topic on the methods for measuring the trace gases both from the in situ and satellite observations. The existing papers are focused only on in situ or satellite measurement methods as well as for only one species (Vairavamurthy et al. 1992; Clerbaux et al. 2003; Cantrell 2008; Martin 2008; Rapson and Dacres 2014; Nemitz et al. 2018; Gonzalez Abad et al. 2019). This is the basic motivation of this study to combine the ground and space born observation methods and instrumentations during last three decades in this topic to make a critical review work for the readers. This review mainly focuses on the fundamental and advanced methods and instrumentations for measuring the fluxes of atmospheric pollutant gases between the earth-atmosphere interaction from the in situ and satellite observations, especially for the air pollutants. This review paper is aimed to figure out the basic principles, advantages, and disadvantages of the methods and instrumentations for measuring the atmospheric pollutants based on the in situ experimental and satellite remote sensing techniques. This paper also makes some recommendations to the readers/researchers for undertaking the right use of the discussed methods. The elaborations of all abbreviations/acronyms used in this study are given in Table S1 (see the supplementary information) for quick reference to the readers.
As the target air pollutants of this paper are the important trace gases of atmosphere, this section describes how to measure the atmospheric trace gases, behaves and influences other energy fluxes in the earth atmosphere interaction. The energy balance between the incoming solar radiation (shortwave radiation) and the outgoing radiation (longwave radiation) emitted from the earth-atmosphere interaction, which can be partitioned into sensible heat flux, latent heat flux, and soil heat flux (Rahman and Zhang 2019c). The energy balance plays a vital role in the evolution of the earth’s climate, and this process is known as radiative forcing (Jennings 1994). This process can be affected by various changing factors such as the changes in atmospheric trace gases (the minor constituents of the atmosphere like CO2, CO, CH4, O3, NO2, N2O, NH3, HCHO etc.), the changes in the overall solar radiation, the changes in the aerosol concentration in the earth-atmosphere interaction. The low concentration (< 1% of all constituents) atmospheric trace gases and their oxidant are highly responsible for the global environmental and climate change aspects. The elemental composition of dry air in volume of the twenty-first century is given in the following Table 1.
In overall observations, the changes in trace gases indicate increasing trend that directly affect the earth’s environment and climate by modifying the radiation budget, and indirectly affected by changing the atmospheric O3 distribution (Wang et al. 1990). Due to the lack of detailed information on atmospheric gases in large scale, there is a research gap in the measurement of atmospheric trace gases and how the global climate affected by the nature and anthropogenic activities. So, the accurate measurement of atmospheric trace gases in local, regional and global scale is very much important.
The methods and instrumentations to determine quantitively and qualitatively the atmospheric trace gases are complex and sometimes difficult due to their low concentration, i.e., the low mixing ratio of the atmospheric trace constituents. The quantity of a trace gas can be expressed in terms of its pressure and density or its mixing ratio (r) which can be written as (Jennings 1994; Wallace and Hobbs 2006):
where, e (e = ρvRvT) is the pressure of the water vapor (the water vapor or any gaseous state that is acting as an ideal gas), p is the pressure of the moist air, M is the molecular weight of trace gas species, and Md is the molecular weight of the dry air. The unit of the trace gas can be expressed as 10^6^ mol/mol ~ ppmV (parts per million volume), 10^9^ mol/mol ~ ppbV (parts per billion volume), 10^12^ mol/mol ~ pptV (parts per trillion volume), µmolm^−2^ s^−1^ (micro mol per m^2^ per second), and nmolm^−2^ s^−1^ (nano mol per m^2^ per second).
There are several methods and instrumentations have been developed during last six decades for measuring the atmospheric trace gases in laboratory, field experiment (in situ) and also from space but there is no cent percentage efficient method to measure trace gases. Every method has its own positive and negative feedbacks that is why this study is aimed to figure out the basic concepts of the existing methods along with their advantages, disadvantages, and also to indicate some important recommendations for the right use of the methods for suitable applications. In the next sections of this paper, we will discuss about the past, present and future possibilities of the potential methods and instrumentations for measuring the atmospheric pollutant gases and oxidants. There are several methods such as colorimetric methods, chromatographic method, eddy covariance method, flux gradient method, mass balance method, spectroscopic methods, and satellite remote sensing methods to determine the air pollutant species and their oxidants in the atmosphere.
Colorimetric method is a sensor-based method, which can determine the molecular compounds such as nitrogen oxides (NOx) (EPA 2017), formaldehyde (HCHO) (Pretto et al. 2000), peroxomonosulfate (SO5^2−^) and peroxodisulfate (SO8^2−^) (Deadman et al. 2017), and raloxifene hydrochloride (Kumar Reddy et al. 2019) by using the coloring reagents. The mostly used colorimetric methods are chromotropic acid method and pararosaniline method as mentioned in the following paragraphs. There are also some other colorimetric methods but they are not well-established.
The chromotropic acid method is a simple and less expensive method for determining the low concentration molecular compounds in indoor and outdoor atmosphere (Fushimi and Miyake 1980; Salas and Singh 1986). A chromotropic acid method is used to sample and quantify formaldehyde (HCHO) (Pretto et al. 2000) with 15.5 µL droplet of chromotropic acid, where the detection rate is limited to 2 ppbV. In the instrumentations, the measurement system is consisting of a sampling chamber, which can be made of Teflon tube with a drop former; a reservoir of chromotropic acid solution that is connected to the drop former and waste tank via a 3-way valve; two optical fibers are connected to the signal processing unit via LED (light-emitting diode) and photodiode; and the signal processing unit is connected to a personal computer. In the experimental process, when a drop of chromotropic acid is done in the sampling chamber, it is occurring violet-pink dye that can be measured by the optical fiber via LED/photodiode sensor as a reference signal. This reference signal is further processed by signal processing unit, and finally acquired by the personal computer. The positive feedback of chromotopic method is that it is easy-care and durable for the quantitative test for formaldehyde in air. The negative feedback is the challenge in the adjustment of P^H^ and reagent blank.
This is a colorimetric method with superior sensitivity, which can be used to determine the less concentration molecular compounds like formaldehyde (HCHO) based on the Schiff reaction (Miksch et al. 1981; Georghiou et al. 1983). The main apparatus of pararosaniline method is the grating spectrophotometers, a capped pathlength optical glass cuvette, temperature controller and the required Reagents (pararosaniline with free base, and pararosaniline hydrochloride) (Kowalska and Jeżewska 2019). The modified method improves the sensitivity and make the operation easy. As trapping medium, the stability of formaldehyde can be improved up to approximately 2 weeks with 10% pararosaniline solution. This method was applied with flow injection method that showed higher sampling rate and easy temperature control (Muñoz et al. 1989). This pararosaniline method is a widely applied method for low concentration measurement. The negative feedback of this method is the susceptibility to interferences.
The chromatographic techniques are working by vaporizing the samples and injecting onto a chromatographic column to detect several carbonyls like formaldehyde simultaneously, which are mainly categorized as gas chromatography (GC) and high-performance liquid chromatography (HPLC) (Ettre 1961; Grosjean and Williams 1992; Falaki 2016). The HPLC method is also applicable for food texting, drug testing, and environmental monitoring (Chen and Wang 2021). A typical gas chromatographic method is consisting of a differential flame ionization injector, and an electron capture detector. The injector temperature should be maintained by a temperature controller at a constant value as required. The column derivatization can be prepared by supports, stationary phase and tubing activities, which provide the separation of peaks, tailing, nonlinearity, and also the reaction of formaldehyde with 2, 4-dinitrophenylhydrazine (DNPH) to produce 1–2, 4-dinitrophenaylhydrazone (HCHO-DNPH), where the hydrazone can be analyzed by GC or HPLC. This technique is widely applied in pharmaceutical sector. The main positive feedback of chromatographic technique is the ability to separate several components at same time along with other facilities like small amount of sample is required. The negative feedback is expensive and need for expert people to operate with care.
In fluorometric methods, three main steps are needed to accomplish the determination of trace i) stripping the trace gas from the gas phase into liquid phase using flow system then apply to the liquid gas analyzer; ii) the reaction of the interested trace gas with the reagent in measurable form; and iii) finally the detection (Lazrus et al. 1988; Trapp and De Serves 1995; Amundson and Zhou 1999). The stripping process can be obtained by various ways based on the fluorescent techniques with different diffusion scrubbers. The main problem in constructing a diffusion scrubber is the clogging of particles on the surface of the membrane that is why the continuous cleaning and/or replacement is needed. However, the stripping solution is recommended to clean the clogging (Lazrus et al. 1988; Heikes 1992). There are two types of technique are used to transfer the gas phase into liquid i) enzymatic technique, which is consisting of oxidation of the trace gas such as the nicotinamide adenine dinucleotide (NAD^+^) is reduced to its fluorescent for formaldehyde (Lazrus et al. 1988); ii) hantzsch reagent technique is a multicomponent reactions for the production of some important oxidant such as 1,4-Dihydropyridines (DHPs). The innovative development of diffusion scrubber is recommended for wide application of this method. Fluorometric method can be used for measuring both the gas and liquid form of species, where the limitation is to construct a diffusion scrubber to remove the clogging.
Eddy-covariance (EC) method is a well-established and widely accepted technique for the measurement of surface-atmosphere flux and gas exchange fluxes with considerable limitations. This method has been utilized for numerous field campaign for several purposes (Mauder et al. 2006; Foken et al. 2010; Rahman et al. 2019). There are thousands of papers discussing the novelty of EC method but almost all the studies noticed some limitations during the measurement and maintenance at site level experiments. Importantly, this method is using continuously for measuring the fluxes of water, energy, and air pollutants in the scientific community of flux measurements responds to the biogeochemical and environmental processes. Basically, this method enable us to insight into the sources and sinks of atmospheric parameters (physical and chemical variables) for testing the process level experiments (Baldocchi et al. 2001; Mauder and Foken 2011; Aubinet et al. 2012). The flux (Fc) of EC method can be defined as the mean average covariance between the deviations from mean vertical wind velocity and the deviations from mean in the interested parameter (e.g., the mixing ratios of trace gases):
where Fc is the flux, w′ is the vertical wind velocity component, c′ is the mixing ratio of the trace gas component, the over bar is presented for the average calculation, the prime denotes the instantaneous deviation from the mean. The basic requirement for EC flux measurement is to measure both the vertical wind velocities and the trace gas of interest at highest possible sample frequency (typically minimum of several Hz). The high sampling frequency will provide the sufficient capture of majority of those eddies that contribute to the atmosphere. The fluctuation of water vapor and temperature is not associated in Eq. (2) that is why it should be corrected (Leuning 2006; Eugster and Merbold 2015) as Eq. (3) for the correct measurement of Fc.
where c¯ is the average value of trace constituent, w¯ is the average value of the vertical wind velocity component, which can be written in terms of water vapor concentration and temperature as-
where cd is the average value of dry air, cv′ is the fluctuated (instantaneous deviation from the mean) value of water vapor. The over bar is presented for the average operation. Finally, the flux density of trace constituent can be written with practical considerations in terms of mixing ratio and flux concentrations as-
However, the EC instruments have been available for more than three decades for the measurements of surface fluxes and atmospheric gases (CO2, H2O and O3), but their performance varies and limits due to the installation, measuring surfaces and seasonal variations (Wilson and Baldocchi 2000; Juráň et al. 2019; Rahman and Zhang 2019b). The measurement of other trace gases like methane (CH4), formaldehyde (HCHO) and nitrous oxide (N2O) were made until 1990, where the performances were vulnerable due to the high noise levels and difficulty in maintenance for the long-term measurement (Nemitz et al. 2018), but the long-term measurement of these trace gases is very important for today’s air pollution and climate change issue. The issue of surface energy imbalance is still under debate, this important issue along with other issues (installation and maintenance) needs more attention of all affected scientific communities to study and discuss the methods for robust correction to remove the uncertainty sources in eddy flux measurements. In EC, the trace gas sensor’s (TGS) cutoff frequency and dynamic time constant should be matched with the measuring site’s atmospheric condition and the covariance between the vertical wind velocity and instrument noise can reduce/limit the flux detection. As per the huge literature, most of the EC flux measurements was performed from single tower over solitary ecosystem which limits the study findings, and recommended to perform more EC measurements over inhospitable environment like arctic tundra, mangroves, inter-tidal wetlands and tropical forests. The regional and global networks of flux towers are established to combine all the measured data in a platform like FLUXNET (www.fluxnet.org) which contains more than 1000 flux towers (Stoy et al. 2013). The main purposes of these flux measurement networks are to convert the data into information and further the information to knowledge using the data assimilation techniques (Rayner 2010). EC method is widely accepted and massively used techniques to measure the surface fluxes and gas exchanges in the atmosphere under some considerations which are indicated as negative feedbacks such as storage terms, impacts of advection, measuring height, and surface heterogeneity.
As per the molecular diffusion, the flux of a trace gas can be measured by the vertical transport of gases in the surface layer –
where K is the constant eddy diffusivity and ∂c∂z is the concentration gradient. The measurement normally performed over a vertical distance, which requires sufficient instruments sensitivity to minimize the minute concentration differences. The overall noise level can be reduced due to the rapid switching between intakes heights perform as high pass filter.
In regardless of eddy diffusivity, the horizontal fluxes can be calculated using mass balance method by determining the balance of horizontal fluxes entering and leaving the desired virtual volume with source (Loubet et al. 2020). The fluxes (Fc) of a trace gas can be expressed as-
where, C is the density of the compound, and U is the horizontal component of wind speed. This mass balance method is widely used to determine the ammonia (NH3) emissions (Yang et al. 2018), and also for measuring other air pollutants like NO2 (Cooper et al. 2017). The main advantage of flux gradient method is the capability of increasing the number of samples per measurement period for achieving the higher resolution on the concentration gradient. The negative feedback of this method is that the above advantage can be achieved only at specific measurement height.
At the time of writing this review paper, most of the fast response instruments for the measurements of trace gases available are based on the range of Laser and optical absorption spectroscopic techniques as described in the following subsection.
Spectroscopy is a basic method for measuring the atmospheric trace gases potentially in the laboratory and the field, which is a technique of differentiating electromagnetic radiation into its material spectrum with optics and solid-state devices (Weidong et al. 2021). In modern science, the spectroscopy utilizes the diffraction gratings for differentiating the electromagnetic radiation, and projected on a charged coupled device (CCD) to extract out the desired spectra. In remote sensing community, spectroscopy method is recognized as a tool which is widely accepted and used to measure, estimate, observe and quantify the gaseous, solid, and liquid in the sections ranging from astronomy to atmospheric chemistry. Spectroscopic measurements are using the absorption features due to specific chemical properties (bonds) for determining the abundance and physical state of the detected species where this technique is also known as imaging spectroscopy, multispectral, hyperspectral, and ultra-spectral spectroscopy (Clark et al. 2003). This technique is used for the measurement of the atmospheric pollutants such as CO, N2O, HCHO, NO2, and SO2. (Yoshii et al. 2003; Harris et al. 2019; Davis and McLaren 2020). This section mainly discussed the spectroscopic techniques for measuring the low weighted molecular compounds and their oxidants. The commonly used and popular spectroscopic techniques are indicated as
The spectroscopic methods and instrumentations are widely used for measuring the atmospheric pollutants, especially air pollutants (NO2, O3, CO, SO2, and particulate matter, PM) and their oxidant products. There are some other spectrometers like Dobson spectrophotometer (Komhyr 1980) and Brewer spectrometer for measurement and validation of ozone (Vaz Peres et al. 2017). Most of the instrumental setup for spectroscopic techniques are commonly used to measure the concentration of the species and expressed in molecules/cm^3^, molecules/cm^2^, mol/mol, and ppm/ppv/ppt. In case of chemically reactive fluxes, the quantity of flux can be measured from the mean concentration measurements which are expressed in µmolm^−2^ s^−1^ or µgm^−2^ s^−1^. A synopsis of some major spectroscopic methods to indicate their positive and negative feedbacks, and some key recommendations to apply a specific method effectively and efficiently for specific purpose is shown in Table 2.
The satellite-based measurement of atmospheric pollutant gases (O3, NO2, CO, SO2, HCHO, and particulate matter, PM) is very much important for its capability to observe the atmospheric parameters in large scale (local, regional and global scale), which enables the scientist to compute the concentration of the atmospheric trace gases for monitoring the air quality and climate. There have been a several numbers of spaceborne/satellite instruments to observe the earth’s atmosphere since 1960 (ESA 2017). Two types of instruments (sensors) are used in the space vehicles (satellites) missions to measure the i) the active instrument used owned artificially generated radiation; ii) the passive instrument used the radiation generated by the sun. Directing principle of instruments also have two i) nadir instrument face directly the earth’s surface, which is preferably used; ii) limb instrument face at the edge of the atmosphere and set measurement at different tangent height. In case of boundary layer applications, the nadir instrument is suitable for atmospheric column at UV–VIS wavelength. The limb instrument is useful to provide the information at vertical resolution and able to retrieve species at higher altitude (Dudhia 2004),
The very beginning of satellite-based practical measurement of atmospheric compounds was started in 1970 using the BUV (Backscattered Ultra-Violet) and TOMS (Total Ozone Monitoring Spectrometer) instruments on Nimbus satellite (Krueger et al. 1973; Heath et al. 1975). After this beginning, there were a steady progress in instrument development for observing the earth’s atmosphere with higher SNR (Signal to Noise Ratio), spectral coverage and resolution as well as the algorithm development. Recently, the space borne observations of air quality and atmospheric chemistry provide the essential contribution to the global observing system (Duncan et al. 2014). In 1995, The GOME (Global Ozone Monitoring Experiment) was launched by ESA (European Space Agency) to obtain the global ozone (O3) along with other atmospheric trace gases (NO2, BrO, SO2, HCHO etc.) within the spectral region of 240–790 nm (UV–VIS-NIR) (Burrows et al. 1997).
In the first decades of 2000s, there have been several number of published potential studies on the satellite estimation of atmospheric trace gases to focus the state of art of the study field and the challenges ahead (Martin et al. 2002; Burrows et al. 2011). Following these studies, the firm progress has been made in the new generation of instruments such as GOME (ESA), OMPS-NM (NASA/NOAA), EPIC (NASA/NOAA), TROPOMI (ESA), EMI (CNSA), TEMPO (NASA), and GEMS (NASA) with enhanced capabilities as well as with higher spatio-temporal resolution. Along with instrument development, there also have been a significant improvement in the retrieval methods, usage of satellite data and the validation of satellite estimation with respect to the in situ measurements. Recently, the satellite based air quality monitoring has become unprecedented due to the near real time (minutes to hour) observation capabilities from the new generation sensors/instruments onboard geostationary satellites like Geostationary Operational Environmental Satellites (GOES-R series, NOAA, NASA), Himawari-8 (on board Japanese Weather Satellite, Japan Meteorological Agency), and Geostationary Environmental Monitoring Spectrometer (GEMS, on board KOMPSAT 2B Satellite, National Institute of Environmental Research, Korea) around the world (Schmit et al. 2017; Kim et al. 2018). This excellent temporal resolution of the geostationary satellite sensors allows the research community of air quality to do the real-time monitoring of different air pollutants. However, this section of this study is aimed to present the basic science of the retrieval techniques and the space borne instruments with solar UV–VIS backscattered radiation measurement and how these instruments have contributed in measuring the atmospheric trace gases for monitoring the surface air pollution and air quality modelling.
The basic concept to estimate the atmospheric trace gases is the measurement of UV–VIS solar backscattered radiation. The satellite remote sensing techniques of trace gases take the advantages of attenuation in the intensity of the backscattered radiation during traversing through the atmosphere, which can be expressed mathematically as Beer’s law-
where Iλ is the attenuated intensity of the backscattered radiation observed by the satellite instrument’s sensor at specific wavelength (λ), I0 is the original intensity of the radiation without any absorption, σλ is the absorption cross section of the trace gas, and Ωs is the slant column profile over the atmospheric path length.The spectral variation in σ to infer Ω are used for trace gas retrieval (Chance 2005; Platt 2006), where the spectral fit is needed to find out the atmospheric abundance over the radiation path, and the radiation calculation is required to determine the path of radiation through the atmosphere. In terms of different air pollutants, the values of σ to infer Ω are changing with respect to the absorption properties of the backscattered radiation from the specific air pollutant as interacted and measured by the satellite sensor. In case of true color images as obtained from satellite sensors, the dark color regions are associated with water/water bodies due to its higher absorption of incoming solar radiation during the interaction; the white color areas are associated with clouds as it reflects most of the incoming energy; the haze (grey/brown) color regions are associated with the air polluting parameters (e.g., smoke, dust, fires, SO2, NO2, O3, CO, etc.) according to the absorption properties of the interacting particles. The radiative transfer method (RTM) is user friendly, and absolutely important tool for several applications in the field of atmospheric physics, climatology and quantitative remote sensing. It (RTM) can simulate the light through a translucent medium in the earth-atmosphere interaction (Seidel et al. 2010). At UV–VIS wavelength, the lower land surface reflectivity is of less than 5% for contributing in molecular scattering to the backscattered radiation (Koelemeijer et al. 2003). In the atmosphere, the instrument sensitivity to the trace gases increases with the increasing reflectivity. The presence of clouds decreases the reflectivity i.e., decreases the instrument sensitivity to the trace gas estimation (Martin et al. 2002; Millet et al. 2006). Optically, all the trace gases can be measured from the space borne observation using the specific spectral region as shown in Fig. 2.
Fig. 2 The optical depth of spectral wavelength for several atmospheric trace gases. Obtained from (Chance 2005)
In nadir geometry, the typical molecules are absorbed in the thin optical wavelengths of 200 nm to 800 nm such as O3 is dominant absorber at shorter than 320 nm and NO2 is dominant over much of 350–450 nm but the commonly conducted retrieval is used between 425 and 450 nm (Martin 2008).
However, the estimation of low weighted trace gases from satellite estimation is highly challenging. The scientists are trying to meet up the challenges by utilizing the recent advent of the availability and applications of machine learning techniques. The satellite estimation of trace gases using solar back scatter and/or thermal estimation is generally driven by an optimal estimation framework (Franco et al. 2018) based on the basic science as indicated above. There are several optimal estimation frameworks for retrieving different trace gases in local and regional scale but the validation is limited due to the lack of ground measurement stations and their high cost (Streets et al. 2013; Wolfe et al. 2019; Zhu et al. 2020). In the operations pathway of retrieving algorithm to obtain the good-quality application-level data (e.g., Level 2 and Level 3 data), the scientists are considering some important conditions such as sky conditions (e.g., cloud fraction should be less than 30%), averaging techniques (time and/or area), row/column anomaly, surface albedo, solar zenith angle, and normalized vegetation index (NDVI). Based on these important factors, the retrieving algorithms are developed by statistical and/or machine learning approaches e.g. the Royal Netherlands Meteorological Institute (KNMI) algorithm, The Dutch-OMI-NO2 (DOMINO v2.0) algorithm, Satellite AErosol Retrieval (SAER) algorithm (Boersma et al. 2011). In some cases, the direct estimation of several satellites is adopted and compared with each other to observe the accuracies (Fu et al. 2019). The main positive feedback of the satellite-based estimation techniques is that it can be utilized from local to global scale and over any type of surfaces, where the negative feedback is the uncertainties in the estimation due to the heterogeneous nature of the earth surface and the seasonal variations in the atmosphere. Most researchers of this field recommended for more research to make available the state of arts of the topic and to enrich the data of estimated and measured results.
The early concepts to determine the atmospheric composition using satellite remote sensing was in nineteenth century. The precise measurement of the infrared through the near UV radiation was obtained by bolometer (Langley 1881), this method was firstly developed by the researchers based on the differential optical absorption spectroscopy for quantifying atmospheric spectra. The sunlight observation was used firstly to estimate the O3 absorption with the aid of photographic spectrometer at different wavelengths (Langley 1902). The preliminary attempt of satellite-based measurement of O3 using backscattered ultraviolet (BUV) sensor with enhanced stratospheric and tropospheric sensitivity on Nimbus-4 satellite in 1970 (Krueger et al. 1973). The Solar BUV (SBUV) instrument was launched on satellite Nimbus-7 in 1979 and the instrument was operated within the range of 255 nm to 340 nm (Fleig et al. 1990). The Total Ozone Monitoring Spectrometer (TOMS) was launched at the same time on same satellite with cross-track scanning capability and global observation at resolution of 50 × 50 km^2^ within the range of 312 nm to 380 nm (Heath et al. 1975). These instruments utilized the Photomultiplier tube (PMT) detector for radiometric measurements at different wavelengths.
The Global Ozone Monitoring Experiment-1 (GOME-1) was launched in 1995 on ESA’s ERS-2 platform to monitor the global ozone within the spectral range of 240–790 nm (Burrows et al. 1997; ESA 2017) along with the particular interest in the space borne estimation of some important tropospheric trace gases like NO2, HCHO, and SO2. The GOME-2 was launched in 2006 on EUMETSAT to obtain the total ozone content and the vertical ozone profile with more high resolution as well as the total column amount of NO2, SO2, water vapor, oxygen/oxygen dimmer, BrO and other trace gases as shown in Fig. 3 (Iapaolo et al. 2007; GOME-2,2021). The Measurement of Pollution in the Troposphere (MOPITT), Multiangle Imaging Spectroradiometer (MISR), Moderate Resolution Imaging Spectrometer (MODIS) were launched on Terra satellite in 2000 for measuring the various atmospheric parameters to monitor the air quality and air pollution in regional and global scale (Deeter et al. 2003; Kaufman et al. 2005; Chen et al. 2008). These instruments were utilizing the upgraded imaging detectors like charge-coupled device (CCD). The MOPITT was equipped for measuring the key compound CO in the 4.7 µm band with 22 × 22 km^2^ resolution and extensively validated with an accuracy and precision of 10% (Edwards et al. 2003). The MISR instrument was covering the spectral bands of 446 nm to 867 nm for retrieving the aerosol abundance and their properties at nine different angles (one at nadir and others at forward and aft-ward view angles) as well as validated with some ground based measuring networks like AERONET (Abdou et al. 2005).
Fig. 3 The optical transmittance of spectral wavelength for several atmospheric trace gases as derived from GOME-2. Adopted from (GOME-2,2021)
The MODIS has the large capabilities for observing the atmospheric parameters with 36 channels and spatial resolution of quarter, half and one km with respect to different channels. The MODIS products are extensively utilized for retrieving different parameters of atmosphere in the energy, water and carbon cycles and validated with several important in situ measurement projects and networks such as BSRN, FLUXNET, GEBA, WATER, and WCRP. (Gilgen and Ohmura 1999; Baldocchi et al. 2001; Rahman and Zhang 2019a; Rahman et al. 2020). The Scanning Imaging Absorption spectroMeter for Atmospheric ChartograpHY (SCIAMACHY) was launched in 2002 (was operational from March 2002 until April 2012) on Envisat satellite for measuring the backscattered solar radiation in both the nadir and limb viewing geometry. It was consisting of 8 channels to measure the spectrum of 240 nm to 2380 nm with typical spatial resolution of 30 × 60 km^2^ (SCIAMACHY 2020). The retrievals of atmospheric trace gases were examined using the SCIAMACHY images (Buchwitz et al. 2005, 2007). The three estimations were also validated with ground-based FTIR measurements with typical bias of 5 to 20% (Dils et al. 2006). It (SCIAMACHY) used the Linear Diode Array (LDA) detector.
The Ozone Monitoring Instrument (OMI) and Tropospheric Emission Spectrometer (TES) was launched on Aura satellite in 2004 to measure the solar backscattered radiation for the purposes of retrieving the atmospheric trace gases like NO2, SO2, BrO, and aerosol characteristics (NASA 2020). The OMI measured the radiation spectrum within 270 nm to 500 nm with spatial resolution of 13 × 25 km^2^ at daily global coverage. The two-dimensional CCD detector was utilized to capture both the atmosphere and earth’s surface with longest record (Levelt et al. 2006). The OMI estimated NO2 was validated with the in situ measurement, and found mean agreement within 30% (Lamsal et al. 2008). For example, the OMI observed NO2, O3, and HCHO over China on May 2021 (monthly averaged) is shown in Fig. 4, which is retrieved based on the commonly used algorithm (Boersma et al. 2011). The Tropospheric Emission Spectrometer (TES) on the same satellite platform measured the spectrum within 3.2 µm to 15.4 µm to observe the tropospheric ozone, reactive nitrogen, and other trace gases in the lowest layer of Earth's atmosphere using FTIR technique (Bowman et al. 2002, 2006). The spatial resolution was of 5 × 8 km^2^. The tropospheric O3 and CO were retrieved by the optimal estimation framework, and also validated with ozonesonde vertical profile network measurements (Worden et al. 2007; OzoneSonde 2020).
Fig. 4 The OMI estimated (monthly averaged) NO
2, O3, and HCHO over China on May 2021(1 Dodson Unit (DU) = 2.6867 × 10^16^ Molecules/cm.^2^)
The Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) instrument was launched on Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) in 2006 with the joint mission of NASA and French Space Agency to observe the vertical distribution of aerosol and cloud properties (CALIPSO 2021). A diode pumped laser (Nd:YAG) was used in CALIOP to detect the backscattered intensity at 1064 nm and the two orthogonal components at 532 nm (parallel and perpendicular)(Martin 2008). The estimation of aerosol from CALIOP was utilized to detect the sources of organic aerosols in the troposphere (Petiot 2012) and validated with the airborne measurement the Cloud Physics Lidar (CPL) data (McGill et al. 2007; CPL 2021). The Infrared Atmospheric Sounding Interferometer (IASI) was launched on MetOP satellite in 2006 to derive the trace gases and cloud cleared radiances in nadir view with typical resolution of 12 × 12 km^2^ (IASI 2021). The HNO3 was derived by a researcher but the all the derivation of trace gases from this instrument products are not validated yet (Wespes et al. 2009).
The Solar Backscattered Ultraviolet Sounder (SBUS) and the Total Ozone Unit (TOU) were launched on FY-3 (A to C) satellite in 2008 by CMA (Chinese Meteorological Administration) to monitor the air quality (SBUS/TOU 2021). The Photo Multiplier Tube (PMT) technique was used in these instruments to measure the radiation within the spectral range of 252 nm to 380 nm for SBUS, and 308 nm to 360 nm for TOU (Huang et al. 2012; Wang et al. 2012). The SBUS ozone profiles provide pretty good precisions at most layers without doing any correction (Fuxiang et al. 2010).
The Ozone Mapping & Profiler Suite (OMPS) on SNPP-NOAA satellite was launched on 2011 to estimate the atmospheric ozone properties in large capacity (higher fidelity with larger swaths), and scheduled to fly on JPSS (Joint Polar Satellite System) to provide the next three decades ozone profile measurements(JPSS 2021). It has three different spectrometers- i) OMPS-Nadir Mapper (OMPS-NM) to observe the total ozone column in the spectral range of 300 nm to 420 nm with 50 × 50 km^2^ spatial resolution (Seftor et al. 2014), ii) OMPS-Nadir Profiler (OMPS-NP) spectrometer to measure the ozone profiles in single pixel in the wavelength of 250 nm to 310 nm with spatial resolution of 250 × 250 km^2^ (Pan et al. 2017), iii) OMPS-Limb sensor (OMPS-L) provides the ozone in the lower stratospheric and tropospheric vertical resolution of 1 km (Seftor et al. 2014).
The Earth Polychromatic Imaging Camera (EPIC) on Deep Space Climate Observatory (DSCOVR)/NOAA satellite platform was launched in 2015 to observe the spectral images within the range of 317 nm -379 nm. The EPIC provides the entire sunlit face of Earth from the LEO (lower earth orbit) with ten (10) channels to offer the retrieval of total column O3 and volcanic SO2. The TROPOspheric Monitoring Instrument (TROPOMI) was launched on Sentinel 5 precursor (S5P) in 2017 to generate the atmospheric composition data (TROPOMI 2021). In recent years, the satellite data products of this instrument have been utilized to retrieve several types of trace gases such as CO, HCHO, CH4, NO2, O3, SO2 etc. with highest spatial resolution of 3.5 × 7 km^2^ within the spectral range of 317 nm to 779 nm (Veefkind et al. 2012). As an example, the global NO2 concentration observed from TROPOMI on 01 May 2021 is shown in Fig. 5, which is retrieved based on the three steps procedure using DOAS method (Van Geffen 2016).
Fig. 5 The global NO
2concentration on 1 May 2021 retrieved from TROPOMI (The values indicated in the boarder of the figure shows the grid coordinates)
The Environmental Monitoring Instrument (EMI) was launched on Gaofen-5 (GF-5) in 2018 by CNSA (China National Space Administration) to monitor the air quality by detecting the pollution trace gases like NO2, SO2, O3, CH2O, and aerosols. The GF-5 is also carrying some other important instrument sensors such as Atmospheric Infrared Ultra spectral (AIUS), Directional Polarization Camera (DPC), and Greenhouse-gases Monitoring Instrument (GMI). The new generation sensors/instruments onboard geostationary satellites like Geostationary Operational Environmental Satellites (GOES-R series, NOAA, NASA), Himawari-8 (on board Japanese Weather Satellite, Japan Meteorological Agency), and Geostationary Environmental Monitoring Spectrometer (GEMS, on board KOMPSAT 2B Satellite, National Institute of Environmental Research, Korea) around the world (Schmit et al. 2017; Kim et al. 2018) are providing high temporal resolution (10 min to 1 h) to observe the instant scenarios of the atmosphere like volcano eruption, lightning, wild fires and dust storms. The summary of above discussed instruments and their main features are listed in the following Table 3.
In the last ten decades, the increasing human activities cause a massive increase in the emission of various pollutant gases of anthropogenic origin, which leads the atmospheric pollution and climate change issue on our living planet. To monitor the atmospheric compositions in different scales, i.e., urban, regional, and global scale, the satellite estimation paved the way where other measurements are rather infrequent. There are several types of applications of space borne instruments in the estimation of stratospheric and tropospheric parameters such as the measurement of meteorological and micrometeorological data, the earth’s surface energy/water balance components (net radiation, sensible heat flux, latent heat flux and soil heat flux), the components of carbon cycle (CO, CO2, CH4, VOCs), the components of the nitrogen cycle (N2O, NO, NO2, O3), and their particles. The retrieval of trace gases measurement from space borne observation is important for monitoring the air quality and climate/climate change analysis.
In general sense, the large-scale estimation of some atmospheric compositions can hypothesize the surface air quality. Sierk et al. 2006 investigated the retrieval and monitoring of atmospheric trace gases using the products of SCIAMACHY. Falke et al. 2001 explored four different examinations of dust and smoke events that affects the air quality using Sea-Viewing Wide Field-of-View Sensor (SeaWiFS) and TOMS satellite data. The mass concentration of aerosol at ground using satellite estimation is mostly focused on the surface air quality research but as well as the concentration of particulate matter (PM) and other trace gases like NO2, O3, and CO. These are becoming more important and prospective due to the decade by decade (1970–2020) developments of the satellite instruments (please see Table 3) for observing the atmospheric chemistry with higher spatial and temporal resolution. Several applications of the products from the satellite instruments to monitor the atmospheric pollutants (PM, NO2, O3, CO, HCHO, and SO2) and to study environmental protection, to conduct epidemiological study as described in the following sub sections.
The primary focus of surface air quality monitoring was the estimation of PM2.5 and PM10 (particulate matter measured at diameter less than 2.5 µm, and10µm, respectively) from satellite and in situ observation based on the spectral signature of the measured radiation. In recent decades, the PM has become an important pollutant that atmospheric researchers tend to focus on, due to the improved capability of retrieving the aerosol optical depth (AOD) using the above listed instruments (please see Table 3) within their spectral ranges. MODIS images were utilized in a study to monitor the air pollution globally with respect to three case studies over Beijing, northern Italy, and Los Angeles, where the results showed that the MODIS derived aerosol optical depth (τa) has a close relation with PM10 (< 10 µm) (Chu et al. 2003). The AERONET measured AOT (τa) and station measured PM10 over Lille (France) were compared along with meteorological variables by means of a semiparametric approach, and found an improved relationship between PM10 and τa (Pelletier et al. 2007). A health risk assessment of PM has been performed in the construction industries and found that majority of the risk assessment performed depending on the old system and neglect the current research findings regarding respiration rate and quantity of the PM studied (Khamraev et al. 2021). Another study on public health effect and economic loss analysis of PM2.5 pollution from coal burning in China, and this study suggests to reduce the coal consumption (Chen et al. 2020).
The current pandemic caused by the severe infections of Covid 19 (Sars.cov-2 virus) around the world is a great threat to human community. Several research works have done and ongoing for mitigating the infection of this dangerous virus, where the influence of PM on the infection of Covid 19 in large scale also investigated by some groups during last two years (2020–2021). A group investigated the global air pollution data and climate data when the Covid 19 cases were greater than 10, and found an association between Covid 19 infections and PM2.5, PM10 with causal link among the PM levels and Covid 19 cases (Solimini et al. 2021). Liu et al. 2021a, b investigated the association between the PM2.5, PM10 and Covid 19 at national and municipal level, and found that the influence of PM2.5 and PM10 (with cutoff point of 35 µg/m^3^) on Covid 19 was more sensitive over England, France, Germany, and Russia but the O3 and PM2.5 were more sensitive over America and Canada, in overall the high concentration of PM2.5 was responsible for spreading the Covid 19 infections. Wang et al. 2020 observed the daily basis PM2.5, PM10 and confirmed cases of Covid 19 over 63 cities of China, and the results indicated that each 10 µg/m^3^ increase in PM2.5, PM10 concentration positively associated with daily confirmed Covid cases, another similar study found every 10 µg/m^3^ increase in PM2.5 and PM10 concentrations affect the Covid 19 case fatality rate (CFR) as 0.24% (0.01–0.48%) and 0.26% (0.00–0.51%), respectively (Yao et al. 2020). Comunian et al. 2020 evaluated the potential influence of PM in spread of Covid 19 via air transmission over Italian cities and found significant results with higher concentration of PM. In these contexts, it can be concluded that the PM is highly responsible for different health hazards along with the infection of novel corona viruses.
Ozone is the first target air pollutant to this scientific community in observing the trace gases from space. A group investigated the long-term change in O3 using various satellite data products (SBUV, TOMS, OMI, GOME-2, SCIAMACHY) to find out the different atmospheric chemical and dynamical factors for the period of 1979–2012. A noticeable correlation was found among the satellite data sets, and the results revealed that the quasi-biennial oscillation dominates the O3 variability at higher latitudes, the total O3 variability was influenced by eddy heat flux as well (Chehade et al. 2014). A review work on the estimation of biomass burning (BB) using satellite data (MODIS, MOPITT) over China along with the field campaign data, showed the important insights into the role of air quality monitoring and living health (Chen et al. 2017).
In case of surface O3, the estimation of it from satellite remote sensing technique is challenging task because the surface O3 is small fraction of total column densities. A background contribution of 20–45 ppbv was observed over USA of O3 concentration in surface air during summer (Fiore et al. 2002). Another group investigated the sensitivity of regional air quality modelling, where they found a significant sensitivity to O3 boundary conditions (Tang et al. 2007). However, the assimilation of data from multispectral sensors and implications of O3 control strategies were indicated by several researchers (Zoogman 2013; Wang et al. 2019).
Generally, the surface O3 is harmful for human health. There are potential studies investigated the association between O3 and health concern (Malmqvist et al. 2014). The present epidemic of Covid-19 also influenced by the O3 concentration. A group of scientists retrieved the O3 from TROPOMI sensor over China during January–February 2020 and validated with FTS (Fourier Transform Spectrometry) observation at Hefei and GPS ozonesonde sensor at Beijing. They found a good agreement with the validation ground data and also the association between O3 and Covid-19 as the higher concentration of O3 might increase the rate of Covid-19 infections and fatality (Zhao et al. 2020). Another study inferred that the high level of ambient O3 and abrupt change in temperature might influences the covid-19 infections over high altitude regions (2500^+^ m ASL) (Semple and Moore 2020). The ozone and its layer are important in the atmosphere in terms of its higher concentration is associated with the infections of novel viruses and the depletion of ozone layer is associated with the increment of ultraviolet rays.
The carbon monoxide (CO) is another important trace gas which have a strong influence in environmental pollution. So, the satellite estimation of CO in large scale is prospective for hypothesizing regional and global air quality. The highest concentrations of CO are explored near the earth’s surface that is why the shortwave infrared measurement of SCIAMACHY and MOPITT are suitable for estimating CO from space which are widely used and validated (Frankenberg et al. 2005; NASA 2019). CO exhibits long-term transport in the atmosphere due to its high life time (~ a few weeks) thus provided the meaningful information on the vertical distribution of CO (Clerbaux et al. 2008). Recently, the TOC of CO was measured from TROPOMI instrument to observe the local enhancement of CO over some cities of Iran (Tehran, Urmia, and Tabriz) using WRF model simulation, and found very good capability of TROPOMI sensor to monitor CO on sub-city scale (Borsdorff et al. 2019). Day by day, the CO concentration is increasing in the atmosphere which effects the greenhouse gases and also affects the human health.
As per USA Environmental Protection Agency (EPA) and other environmental agencies around the world, the NO2 is designated as an important pollutant criterion due to its negative implications on human health, crop yields and climate. Some previous studies on the epidemiology indicated that the exposure of NO2 at moderate and high level caused bronchitis and asthmatics in children, i.e., it reduced the lung function (Burnett et al. 2004; EPA 2020). The abundance of stratospheric and tropospheric NO2 is measured from the space borne sensors within the spectral range of 330 nm to 500 nm based on the direct fit and DOAS algorithm (Bucsela et al. 2013; Xu et al. 2019). The global NO2 (stratospheric, tropospheric, and total column densities) was estimated from OMI sensor based on the DOAS method showed the evidence of distinct development over urban and industrial regions (Bucsela et al. 2006). The space borne observation of tropospheric NO2 from GOME and OMI sensors were retrieved and correlated significantly with the ground based NO2 measurements (Lamsal et al. 2008). Uncertainty analysis in satellite estimation is very common, accordingly the satellite estimation of NO2 is also inclined to uncertainties. (Zara et al. 2018) investigated the NO2 estimation from OMI and GOME-2 sensors, where the SCD (slant column densities) uncertainties were increased by 1 to 2% per year for OMI and by 7 to 9% per year for GOME-2.
Very recently, the regional and global scale analysis of NO2 from space were focused on the association between NO2 concentration and Covid-19 outbreak. (Zhang et al. 2021) examined the global association among NO2 (retrieved from Sentinel 5P), Covid-19 pandemic, and lockdown policies; the results indicated a thought of heterogeneous patterns of associations among them. Ogen 2020 used the Sentinel 5P data for mapping the tropospheric NO2 over 66 administrative regions in Spain, Germany and France during Covid 19 outbreak, where the results showed that the chronic exposure to the NO2 pollutant could be a vital contributor to the high rate of Covid-19 fatality. Recent studies indicate that the higher exposure to NO2 is bad for human health and infectious to novel viruses.
The primary sources of SO2 are the anthropogenic sources such as burnings of fossil fuels, smelting of sulfur-containing elements, and the natural sources are oxidation of dimethyl sulfide from marine phytoplankton, volcanic activity, and negligible from soil and vegetation decay. The massive volcanic eruption of SO2 produces sulfate aerosol which might persist in the atmosphere for about of one year thus severely affect the environment (McCormick et al. 1995). There is significant evidence of association between SO2 (its sulfate) and public health (WHO 2016). As the local- to global-scale estimation of SO2 can contribute to the research of air quality monitoring and climate issues, the satellite estimation of it is highly needed but the observation of SO2 from space is difficult due to some factors. Firstly, the ultraviolet (UV) region of SO2 is merged with strong absorber of O3; secondly, it lies closely near the surface due to its short-life time i.e. the measuring layer is located where the satellite sensitivity is too low; thirdly, the features of SO2 absorption are weak in nature (Veefkind et al. 2007). These difficulties increases the uncertainties in the satellite estimation of SO2, still the estimation approaches are going on (Fioletov et al. 2017). The continuous development of hyperspectral measurement sensors (e.g., GOME-1/2, TOMS, OMI, SCIAMACHY, OMPS etc.) on different satellite platforms enable us to observe the SO2 during various anthropogenic and natural events (Krueger et al. 2000; Yang et al. 2013; Akyuz and Kaynak 2019). A group of scientists estimated the SO2 from OMI sensor over eastern USA, eastern Europe, northern China, and India during 2005 to 2015 (Krotkov et al. 2016). They found the declining trend over eastern USA, eastern Europe during 2005 to 2015 and also the declining trend over China (Elissavet Koukouli et al. 2018) during 2011 to 2015 due to their trend to use green technologies and strict regulations settled by the governments of those countries but the increasing trend was observed over Asian cities which showed the fate of developing and underdeveloped countries. This declining trend over USA and China can teach the Asian regions to implement the strict regulations on the emissions from industries, transportations and other sectors.
Basically, HCHO is generated in the environment due to the oxidation of CH4 and non-methane VOCs (Lee et al. 2005) along with the oxidation of VOCs from biomass burning, on road vehicles and industries (Zhu et al. 2016). The first satellite estimation of HCHO was performed using GOME-1 data over north America, and demonstrated the ability to measure the HCHO from space and also provided meaningful information of the emission of reactive hydrocarbons (Chance et al. 2000). (Palmer et al. 2001) determined the HCHO from OMI data using a direct fit of the HCHO absorption by converting the SCD to VCD. After this beginning, several satellite estimations were performed using the data from different sensors such as SCIAMACHY, OMI, GOME-2, SNPP, NM-OMPS, and TROPOMI to retrieve HCHO over various areas around the world (Kurosu et al. 2004; Zhu et al. 2016; De Smedt et al. 2018). The satellite estimated HCHO was used for providing the information on the oxidation capability of the atmosphere with global change in biogenic emissions (Valin et al. 2016; Jin et al. 2017). The seasonal change in the biogenic emissions was observed using the GOME-2 with HCHO estimation over southern USA during May 1996 to September 2001, and found 25% higher emission at beginning and lower at the end of growing season (Palmer et al. 2006). Another potential satellite estimation of HCHO in terms of necessity of biogenic, pyrogenic, and anthropogenic hydrocarbon emissions was performed over Asia and some other important cities worldwide during 1997 to 2009 (De Smedt et al. 2010). They found that the HCHO column was increased over China by 4% per year and 1.6% per year over India, whereas it was decreased by 3% per year over Tokyo and cities of northern America. However, the past and present scientific application of satellite retrieval of HCHO showed a good understanding but still need to address further issues like the optical depth of HCHO.
Based on the studies and discussions in the previous sections, this study highlighted some recommendations to the readers/researchers so that they can focus on the specific issues of the methods and instrumentations for measuring the atmospheric pollutants for different applications from in situ and satellite observations, and also to find out the research gaps in this topic for further prospective researches. This study pointed out some specific recommendations as indicated in the following sub-sections.
The basic concept to measure the trace gases from the space to differentiate the different wavelength regions, which are sensitive to various vertical atmospheric regions. In the boundary layer, this differentiation is used to discriminate the free troposphere, where a satellite instrument measures a trace gas simultaneously at different wavelengths such as CO at 4.7 µm region (sensitive to free troposphere) and at 2.4 µm region (sensitive to entire troposphere). The separation of the backscatter region (2.4 µm) is considerably contented to determine CO in the troposphere than that of at the thermal region (4.7 µm) in the free troposphere (Pan et al. 1995). Similar technique has been utilized to observe the O3 profiles which showed the increased ability to separate O3 in the boundary layer (Di Noia et al. 2013). The atmospheric reflectance is affected by the solar zenith angle hence it is an important satellite design parameter, where the minimum recommended values is > 70º. Changes in radiance are also crucial factor for designing satellite instruments e.g. the presence of land surface may increase the satellite radiance by 4 to 9%, and 10 to 20% for the presence of sea surface (Aas 1995). The further development and application of these techniques is required in the design of satellite sensors.
Cloud is an obstacle for observing the atmosphere from space. Cloud free observation is highly needed for more accurate estimation of atmospheric parameters from space borne observations. Several techniques are utilized to obtain the cloud free observations, where the resolution of satellite sensors is an important factor. The higher resolution of satellite sensor can provide the more cloud free observations. Some potential studies, it was found that the higher spatial resolution could provide the higher number of cloud free observations from satellite (Krijger et al. 2007; Li et al. 2020). The revisiting capability of a satellite sensor is also an another key factor that influences the ability to acquire the cloud free images (Ogunbadewa 2012; Wu et al. 2021). The minute to hour temporal resolution of geostationary satellite sensors allow to do the diurnal analysis for understanding the chemical transformation and/or atmospheric turbulence throughout the day. This feature of geostationary satellites also increases the probability of cloud free observation by the high revisiting capability (solve the coinciding with cloud problem) of every 10 min. Currently, the geostationary orbiting sensors like GOES-R series over North America, Himawari-8 over East Asia and Western Pacific, and GEMS over Asia’s data are available. Therefore, the satellite instrument design scientists are focuses on these factors and the data users of the satellite sensors also pay attention to it.
Data assimilation is the process of adding values to the observations by filling in the gaps of observations (satellite or in situ) and/or adding values to the models by forcing the measured values by means of mathematical techniques. Data assimilation of several satellite instruments and ground based measurements are very much important for different issues such as weather and air quality forecasts, evaluation of methods, models and instruments, assessment of relative values and added values to Global Observing System (GOS) (O’Neill et al. 2004; Lahoz and Schneider 2014). In air quality modeling, a previous study performed the assimilation of 0–6 km O3 column from MetOp (Meteorological Operational satellite) and found that it improve the quality of O3 forecasts at regional scale (Boisgontier et al. 2008). Werner et al. 2019 investigated the assimilation of surface PM2.5 and satellite AOD data with WRF-Chem model over Eastern Europe, and found that the model results were highly influenced by the assimilation of satellite data.
Validation of satellite estimation is crucial for any development and application-based study, where the validation for trace gas measurements is challenging due to the lack of frequent and large-scale in situ measurements and their networks. Recent developments in the in situ instrumentations and their field campaigns paved the ways to perform some potential studies on validation of different trace gases. The validation efforts for HCHO estimation from OMI sensor with respect to in situ observations with respect to in situ observations (12 aircraft campaigns) over large spatio-temporal area and suggested for the improvements in the correction of systematic biases of OMI HCHO product (Zhu et al. 2020). The NO2 pollutant was estimated from TROPOMI and validated with ground measurements near the Canadian Oil sands, and found an excellent agreement (Griffin et al. 2019). The total column ozone was retrieved from two instruments e.g. INSAT-3D (India) and Atmospheric Infrared Sounder (AIRS), and compared with the in situ measurements (Dobson Spectrophotometer) over India, and obtained a very good agreement (Kumar et al. 2021). They also investigated the seasonal variations and obtained the RMSE as 11 to 16 (0.68 to 0.80), where the higher value was obtained during pre-monsoon and lower value was obtained during winter season. As per the conclusions of the past and present studies on validation of satellite estimated trace gases, further developments in satellite and ground measurements are needed along with continuing study on this issue.
Basically, the satellite retrieval algorithm for trace gases depends on the external information of geophysical properties such as the surface reflectivity, surface emissivity, and species profiles which influences the solar backscattered radiation (UV, NIR, TIR). If the desired measuring surface/partial surface is covered with snow or ice then there will be large amount of uncertainty in satellite estimation of trace gases (Martin et al. 2002; Boersma et al. 2004). The cloudy pixels and their adjacent pixels added more error bias in aerosol estimation (Wen et al. 2007; Marshak et al. 2008), whereas the specific gas profiles and aerosol properties at local scale can reduce the error bias (Wang and Martin 2007). The formation of boundary layer (morning time, day time, evening, and night time) also influenced the retrieval of trace gases that is why it was suggested to consider it in the retrieval algorithm and model development (Bösch et al. 2018).
The quality control of data should be in account with retrieval algorithm and model development. A potential study showed that a post-screening and bias correction method was included with the IAPCAS-GOSAT (Institute of Atmospheric Physics Carbon Dioxide Retrieval Algorithm for Satellite Remote Sensing-Greenhouse gases Observing SATellite) retrieval techniques, which reduced 65 to 70% of the data after quality control and significantly improved the error reduction as of 84% (Yang et al. 2017). Machine learning (neural network, support vector machine, decision tree, random forest etc.) has become important and suitable technique due to the availability of new and high resolution datasets to predict the trace gases in the atmosphere at local to global scale, and analyzed them in terms of multivariate, nonlinear, non-parametric regression or classification analysis (Lary et al. 2016). Further development in computationally efficient optical estimation (OE) methods is required to determine the model sensitivity to different parameters (Sandu et al. 2005; Nguyen et al. 2019). Several statistical modeling techniques can be suggested in the way of algorithm and model development such as the geographically weighted regression modeling (GWR) is highly recommended for the spatial observation with per pixel information over the area of interests (Rahman et al. 2022).
In recent years, the satellite instrumentation is in the advanced developed stage for monitoring the atmospheric trace gases and their particles from space. The main development is happened in higher spatial and temporal resolution, revisiting capability, angular viewing capacity, and the long-term operation period. Most of the meteorological satellites are sun-synchronous but the multiple satellite instruments on different orbits are required for diurnal analysis. This important feature can be achieved by developing the geostationary mission to ensure the continuous observations. The features of a newly launched instrument called Geostationary Environment & Monitoring Spectrometer (GEMS) in 2020 on the GEOstationary Korea Multi-purpose Satellite-2 (GEO-KOMPSAT-2) for observing the hourly air quality and pollution over Korean peninsula and Asia Pacific regions (Kim et al. 2020). This mission also hosts two other i) Advanced Meteorological Imager (AMI), and ii) Geostationary Ocean Color Imager -2 (GOCI-2). The two other geostationary satellite TEMPO (Tropospheric Emissions: Monitoring of Pollution) and Sentinel-4 are almost ready to lunch in near future to complete the earth constellation for observing hourly the whole earth along with the GEMS (Chance et al. 2013; ESA 2022). The Geostationary Operational Environmental Satellite (GOES-R) R series is adding some new sensors like Extreme Ultraviolet and X-Ray Irradiance Sensor (EXIS) and Solar Ultraviolet Imager (SUVI) to detect the solar flares, coronal holes, extreme ultraviolet rays which impact the upper atmosphere; Space Environment In Situ Suite (SEISS) and Goddard Magnetometer (GMAG) to monitor the energetic particles and the magnetic field dynamics which are responsible for the radiation hazards (NOAA-NASA 2021). Some operating instruments on board different satellite platforms are going to end, i.e., the EOL (expected end of life) is in near future like OMI, TES, and AIRS on board Aura (NASA) whose EOL is about 2023. The long-term operation and continuation of a mission with enhanced capability is required to confirm the continuous study on that instrument/satellite mission for any desired application. That is why some of the important previous missions are recommended to join with future missions such as OMPS-NM, LM (NASA) scheduled to fly on JPSS (Joint Polar Satellite System) to provide the next three decades’ ozone profile measurements with three different spectrometers. Therefore, the potential evaluation and efficient applications of the present instruments and the supports to the next generation instruments and their satellite platforms are highly recommended.
The adverse effects of air pollution on human health are studied in terms of different health problems such as respiratory, cardiovascular, metabolic, neurological, and genitical concern to evaluate and analyze the epidemiological and toxicological data. Several organizations (ATS, EEA, EPA, ERS, HEI, ISEE, WHO) around the world consider the air pollution and human health issue, and settled some standard values on different pollutants (PM, NO2, O3, SO2 etc.) and their particles based on the latest science on air pollution and epidemiology (Hoffmann et al. 2020). The latest combined meeting/workshop (21–22 January 2020, Brussels, Belgium) of HEI, ERS, WHO, ISSE to follow-up the latest science, studies and their results and recommendations, and focused on the following points and suggestions.
However, these points and suggestions are very much important for any further research on air pollution and epidemiological study to make some new standards, policies and regulations.
The epidemic of Covid-19 has intensively affected our planet and has caused about 4 million deaths globally (as on July 2021) since December 2019 which is a great threat to public health. The ambient air pollution (PM, NO2, O3, SO2 etc.) is associated with human health problems, especially with the respiratory related problems like asthma, bronchitis etc. Covid-19 is a very strong virus of Severe Acute Respiratory Syndrome (SARS) group that is why the association between ambient air pollution and Covid-19 infections and fatality is highly possible. The study approaches for investigating the health impact of ambient air pollution on pandemic diseases are different from the chronic diseases, where both the cohort and ecological studies are taken into consideration (Villeneuve and Goldberg 2020). There was previous record of pandemic due to SARS coronavirus in November 2002 but not in severe like current epidemic, where the association of ambient air pollution and SARS coronavirus fatality was also evident (Cui et al. 2003). The clinical study of Covid-19 indicated the smoking contribution to the epidemic and the mechanistic process also showed that the air pollution may influenced the involvement of angiotensin converting enzyme 2 receptor in Covid-19 epidemic (Aztatzi-Aguilar et al. 2015; Guan et al. 2020; Heederik et al. 2020). Based on the findings and discussions of the literature reviewed (please see “Satellite estimation techniques” section) on the association between ambient air pollution and Covid-19, this study can comment the following points.
This paper reviews the previous and recent studies on the measurement techniques of atmospheric pollutants (PM, O3, CO, NO2, SO2, and HCHO) from ground and satellite observations. It summarizes the main scientific concepts, positive and negative feedbacks of the methods and also provides some recommendations regarding the prospective research gaps in this research field. In a broad sense, there has been a well-founded progress but the research still faces many challenges. As the air pollution is an important health issue, the several national and international organizations (EPA, EEA, HEI, WHO etc.) are trying to make the bridge between air pollution measurement science and epidemiological study to update with the recent status of air quality around the world. The in situ–based measurement methods like colorimetric, chromatographic, eddy-covariance, flux gradient, spectroscopic etc. are discussed in this study and found that each method has its own merits and demerits relative to others. The is no consensus on which one is the best method but the eddy-covariance and spectroscopy methods are commonly utilized and validated in the laboratory and field. However, the recent advancement in electronics, computer and communication engineering made the ground based measurement, data collection and processing more faithful, efficient, and user friendly. But the lack (due to the highly expensive and unreachable measuring places over the heterogeneous land surfaces and deep ocean) of sufficient regional and global coverage make restriction to achieve the unified and world-wide in situ measurement networks. To achieve the regional and global scale coverage of air pollutant measurements, the space-borne measurement is utilizing snice 1960 and it has been developed remarkably by this century. This paper narrates the developing history of space borne instrumentations by focusing their scientific techniques and capabilities to measure the gaseous and other air pollutants over several parts of the world as missioned by several space administrations like NASA, CNSA, ESA, JAXA etc. The applications of space-borne instruments such as MODIS, OMI, GOME, SCIAMACHY, MOPITT, TROPOMI, GEMS etc. are also discussed in terms of the air pollutant measurements. Finally, this study makes some recommendations to the researchers by highlighting- satellite design parameters which are playing important role in obtaining the cloud free observations and high resolution imageries; data assimilation and validation for fusion of in situ and satellite measurements; algorithm and model development to estimate air pollutants in more extent for different applications; future prospect of satellite instruments which is very much important to this research community to generate novel idea for upcoming research; and the epidemiological study and policy making on public health issues to enhance and implement the air pollution control policies in this regard.
Below is the link to the electronic supplementary material.
This study wishes to acknowledge the intellectual contribution of the department of EEE, Pabna University of Science and Technology, Bangladesh.
The author declares no competing interests.