# Why satellites?

Extracted from Chimot, J., Global mapping of atmospheric composition from space – Retrieving aerosol height and tropospheric NO2 from OMI, PhD book, Delft University of Technology (TU Delft), The Royal Netherlands Meteorological Institute (KNMI), July 2018.

Trace gas and aerosol can be observed from a series of ground-based (in situ or re- mote) sensors. For instance, the global Aerosol Robotic network (AERONET) consists of several hundreds of ground-based carefully calibrated Sun photometers that mea- sure extinction aerosol properties at seven wavelengths at several locations in the world (Holben et al., 1998). Their high accuracy is a great asset for satellite aerosol product calibrations and validations (Lee et al., 2012; Levy et al., 2013; Sayer et al., 2013; Jethva et al., 2014). MAX-DOAS technique allows to provide between 2 and 4 pieces of information on tropospheric NO2 – Nitrogen dioxide, SO2 – Sulphur dioxide and HCHO – Formaldehyde profiles, such as in China (Vlemmix et al., 2015; Wang et al., 2017). Other networks like PANDORA or ACTRIS are also of interest for trace gas observations. Ground-based instruments have several advantages:

• Related observations are usually very accurate.

• A lot of measurements contain information about concentrations at the the altitude where we breathe.

• Usually, time series are acquired with a very high frequency, thus providing insight in the temporal evolution.

However, exploitation of such instruments is hampered by a series of drawbacks:

• Measurements are not only sensitive to the local emissions but also weather conditions (i.e. the spatial extension of the gas and particle plume is driven by the wind direction). Therefore, they may not always be representative of the local implemented policies or activities. In the context of flux estimates, repre- sentativity errors are highly caused by the variability of surface concentrations (mostly governed by local conditions) and vary with respect to the station type (oceanic, coastal or continental). Furthermore, over a complex source region such as the Seoul metropolitan area (Republic of Korea), it is essential to have a high-resolution observing system with a large enough coverage to be able to differentiate changes among individual sources (such as NOx ) (Duncan et al., 2016). Ground-based sensors are usually more sensitive to their direct envi- ronment but not, or little, sensitive for a large area.

• Ground-based networks are usually limited by the spatial coverage: i.e. within a limited area close to their location. Therefore, they do not offer a full global representativity. For example, AERONET coverage remains somehow irregular: very dense over most of lands, less in Russian forests and Central Africa, and very little over ocean and narrow seas. Deployment of MAX-DOAS in China is very limited, usually restricted to specific research studies. Not all of the surface stations are properly located to determine concentration changes (such as NO2 – Nitrogen dioxide or CO2 – Carbon dioxide) at all times (Ciais et al., 2010). Other sources, such as ship emissions, are difficult to monitor from ground. Finally, emission and sink information from politically closed countries (e.g. North-Korea) are not accessible.

• Each instrument may include a different calibration protocol leading to inhomogeneities in the time series, or between measurement sites.

• Observations require a measurement system that can sample (in daytime) up to the free troposphere, not just the affected surface layer, to avoid undue in- fluence of local natural signals and to obtain regionally representative mea- surements. For example, NO2 ground-based lidars (e.g. RIVM) are mostly well validated in the boundary layer and highly sensitive to its formation and vari- ation, but much less to higher altitude (Volten et al., 2009; Piters et al., 2012)

The need of satellite data, in combination with ground-based networks, is crucial for global mapping of aerosol particles and trace gases in the troposphere (Burrows et al., 2011). With their global coverage and high spatial and temporal resolutions, satellite technology provides the unique potential to move from data-poor conditions toward a data-rich system. Several remote sensors have demonstrated their ability to provide relevant information on changing air quality (Richter et al., 2005; Martin, 2008; Boersma et al., 2011; Levelt et al., 2017), long-range transport of pollutants (Edwards et al., 2006), infer emission locations and strengths (Clerbaux et al., 2009; Frankenberg et al., 2008; Streets et al., 2013; Ding et al., 2015; Levelt et al., 2017; Fioletov et al., 2017), O3 – Ozone hole (Levelt et al., 2017), and climate change issues (Worden et al., 2008). Furthermore, the relative long life-time of satellite instruments (typically more than 5 years) fulfill the need of trend analyses which require continuous and long-term observations to understand societal changes (e.g. economic crisis, technology implementations) (Boersma et al., 2007; Castellanos and Boersma, 2012; Duncan et al., 2014, 2016).

The figure below is an illustration of the possibility of satellite sensors (e.g. GOME-2) observing long-range transport of particles (Saharan dust and absorbing black carbon released by forest fires in Portugal in October 2017).