Based FTIR Spectra for the Region of Thessaloniki and Validation with TROPOMI

Laboratory of Atmospheric Physics – Aristotle University of Thessaloniki

Measurements of full atmospheric gas columns ( , , , , ) using ground- based FTIR spectra for the region of Thessaloniki and validation with TROPOMI

Author :

Mermigkas Marios Supervisor :

Professor Balis Dimitrios

ΑΡΙ΢ΣΟΣΕΛΕΙΟ ΠΑΝΕΠΙ΢ΣΗΜΙΟ ΘΕ΢΢ΑΛΟΝΙΚΗ΢

΢ΧΟΛΗ ΘΕΣΙΚΩΝ ΕΠΙ΢ΣΗΜΩΝ

ΣΜΗΜΑ ΦΤ΢ΙΚΗ΢

Π.Μ.΢. ΦΤ΢ΙΚΗ΢ ΠΕΡΙΒΑΛΛΟΝΣΟ΢

Μεηρήζεις καηακόρσθης ζηήλης αημοζθαιρικών αερίων ( , , , , ) με ηη τρήζη θαζμαηοζκοπίας σπερύθροσ μέζω μεηαζτημαηιζμού Fourier για ηην περιοτή ηης Θεζζαλονίκης και διαδικαζία επικύρωζης με ηο δορσθορικό όργανο TROPOMI

Μέρμηγκας Μάριος

Επιβλέπων Καθηγητής : Μπαλής Δημήτριος

ACKNOWLEDGEMENTS

First of all I would like to express my appreciation to Professor Dr. Dimitrios Balis for his valuable help and for the time he spent in order to help me get through this study as well as for the fact that he inspired and encouraged me through all these years.

Also , we are grateful to KIT (Karlsruhe Institute of Technology) for providing us with the instrument (EM27/SUN) , which was developed by Bruker Optics and of course Professor Frank Hase and his team for the collaboration.

In addition , we want to thank the Department of Meteorology and Climatology ( AUTH ) for the meteorological data.

Laboratory Teaching Stuff Chrysanthi Topaloglou is gratefully acknowledged for her precious help and for her instructions during all these months. Furthermore , I want to thank Post-doctoral Research Associate Mariliza Koukouli , PhD students Voudouri Kalliopi-Artemis and Konstantinos Michailidis and Post-doctoral Researcher Nikolaos Siomos for their important help.

Finally , special thanks to my mother , my father and my brother for their support and the sacrifices that they have done for me.

Table of Contents

ABSTRACT ...... 11 ΠΕΡΙΛΗΨΗ ...... 13 1.) Introduction ...... 16 1.1) Climate change ...... 16 1.2) The greenhouse effect ...... 21 1.3) The anthropogenic Green-House Effect ...... 24 1.4) Changes in radiative forcing ...... 26 2) Greenhouse gases ...... 27 2.1) Carbon dioxide ...... 28

2.2) Methane (CH4) ...... 30 2.3) Water vapour (H2O) ...... 32 2.4) Carbon monoxide (CO) ...... 35 2.4.1) Human Impact ...... 35 2.4.2) Potential for control ...... 36 3.1) Interaction of radiation and matter ...... 37 3.1.1) Molecular Spectroscopy ...... 38 3.1.2) Molecular vibrations ...... 38 3.1.3) Molecular Potential ...... 39 3.2 ) Discovery of Infrared light ...... 45 3.3 ) What is an Infrared spectrum ? ...... 47 3.4 ) What kind of molecules absorb infrared light ? ...... 49 4) The Fourier Transform Principal ...... 50 4.1) EM/27 SUN BRUKER INSTRUMENT ...... 54

4.2) Theory of EM27/SUN - BRUKER ...... 58 4.3) Utility and advantages ...... 63 4.4) What is an FTIR spectrum used for ? ...... 63 4.5) Interferometer and Fourier Transform ...... 67 4.5.1) Michelson’s Interferometer...... 67 4.5.2) Interference Pattern ...... 69 4.5.3) Michelson’s Interferometer principle ...... 71 4.6) Fourier Transform of Interferogram to Spectrum ...... 74 4.7) Extracting the spectrum from raw data ...... 76 4.8) The Fast Fourier Transform (FFT) ...... 78 4.9) Background Spectrum ...... 79 4.10) Absorption spectroscopy ...... 81 4.11) Transmission spectrum ...... 81 5) Analysis of FTIR ground-based measurements ...... 89 5.1) CamTracker program ...... 89 5.2) Recording Spectra ( OPUS_7.2.139.1294 ) ...... 93 5.3) TCCON ...... 94 5.4) The HITRAN Database ...... 95 5.5) COCCON software description, version: 180806 ...... 98 5.5.1) Process ...... 99 5.5.2) Starting the retrieval ...... 102 6) Misalignments ...... 105 6.1) Ghost to parent ratio ...... 106 6.2) Pressure broadening ...... 107 6.3) Doppler broadening ...... 107 6.4) Air – mass dependency ...... 108 6.5) ILS ...... 109 6.6) Zero filling ...... 111

6.7) Apodization Function ...... 112 6.8) Phase Error ...... 114 7) Ground-based FTIR measurements in Thessaloniki ...... 116 7.1) Set up ...... 116 7.2 ) Mean daily values and daily course of FTIR ground-based measurements in Thessaloniki ...... 118 7.3 ) Standard deviation of x-gases ( Daily values ) ...... 135 7.4 ) Comparison with other sites ...... 140 8) Validation with TROPOMI satellite sensor ( Sentinel - S5P ) ...... 148 8.1) Sentinel 5 – Precursor ...... 150 8.2) EM27/SUN validation with TROPOMI data products (XCO,XCH4)… ...... 155 9) Conclusions ...... 164 References……………………………………………………………………………………….166

ABSTRACT

In this master thesis , our aim is to extract the daily course of greenhouse gases ( such as carbon dioxide ( ) and methane ( ) ) and also study the effect of greenhouse gases in the atmosphere through Fourier Transform Infrared Spectroscopy by measuring the full atmospheric gas column from the ground where the system (EM27/SUN) is faced up to the space and also to interpret the results for the region of Thessaloniki. The EM27/SUN – FTIR was place up on the roof of the Laboratory of Atmospheric Physics (AUTH) and our measurements were started from 15th of January in the morning.

The EM27/SUN spectrometer is intended to measure direct solar radiation in the near infrared (NIR) spectral range. The recorded spectra contain signatures of atmospheric constituents ( , , , , ), which can be evaluated to retrieve the total columns. The system was built and optimized explicitly for analysing atmospheric gases by using the radiation of the sun as a light source for high precision gas analysis and was developed by KIT in collaboration with Bruker Optics.

We managed to take our first total column measurements of X-GHGs in the area of Thessaloniki using FTIR spectroscopy for a period of 8 months (January – August 2019) , observe diurnal and seasonal variability of X columns of GHGs and investigate possible sources and correlation with meteorological data.

Finally , validation with Sentinel – 5 Precursor satellite sensor , TROPOMI , was achieved for data products such as carbon monoxide and methane exporting a correlation factor between satellite data and ground-based FTIR measurements. Furthermore , we tried to compare our results so far with measurements from different European and world wide sites.

ΠΕΡΙΛΗΨΗ

΢τθν παροφςα διπλωματικι εργαςία ο ςτόχοσ μασ είναι να εξαγάγουμε

τθν κακθμερινι πορεία των αερίων του κερμοκθπίου (όπωσ και ) και να μελετιςουμε επίςθσ τθν επίδραςθ των αερίων του κερμοκθπίου ςτθν ατμόςφαιρα με φαςματοςκοπία υπερφκρου μζςω μεταςχθματιςμοφ Fourier μετρϊντασ τθν κατακόρυφθ ατμοςφαιρικι ςτιλθ των αερίων από το ζδαφοσ όπου βρίςκεται το ςφςτθμα (EM27 / SUN - BRUKER) μζχρι το διάςτθμα και να ερμθνεφςουμε τα αποτελζςματα για τθν περιοχι τθσ Θεςςαλονίκθσ.Σο φαςματόμετρο τοποκετικθκε ςτθν ταράτςα του εργαςτθρίου Φυςικισ Ατμόςφαιρασ του Α.Π.Θ. και οι μετριςεισ μασ ξεκίνθςαν το πρωί τθσ 15θσ Ιανουαρίου.

Σο φαςματόμετρο EM27 / SUN προορίηεται για τθ μζτρθςθ τθσ άμεςθσ θλιακισ ακτινοβολίασ ςτθ φαςματικι περιοχι ςτο κοντινό υπζρυκρο (NIR). Σα καταγεγραμμζνα φάςματα περιζχουν υπογραφζσ των

ατμοςφαιρικϊν ςυςτατικϊν (XH2O, XCO2, XCO, XCH4, XO2), τα οποία μποροφν να αξιολογθκοφν για τθν ανάκτθςθ των κατακόρυφων ςτθλϊν. Σο ςφςτθμα καταςκευάςτθκε και βελτιςτοποιικθκε αποκλειςτικά για τθν ανάλυςθ των ατμοςφαιρικϊν αερίων χρθςιμοποιϊντασ τθν ακτινοβολία του ιλιου ωσ πθγι φωτόσ για ανάλυςθ αερίων, υψθλισ ακρίβειασ και αναπτφχκθκε από το πανεπιςτιμιο τθσ Καρλςροφθσ ςε ςυνεργαςία με τθν εταιρία Bruker Optics.

Καταφζραμε να πάρουμε τισ πρϊτεσ μασ μετριςεισ κατακόρυφων ςτθλϊν των ατμοςφαιρικϊν αερίων ςτθ Θεςςαλονίκθ χρθςιμοποιϊντασ φαςματοςκοπία υπερφκρου για μία περίοδο 8 μθνϊν ( Ιανουάριοσ – Αφγουςτοσ 2019 ) , να παρατθριςουμε τθν θμεριςια και εποχιακι μεταβλθτότθτα των ςτθλϊν των αερίων του κερμοκθπίου και να διερευνφςουμε πικανζσ πθγζσ ςε ςυςχετιςμό με μετεωρολογικά δεδομζνα.

Σζλοσ , θ επικφρωςθ με τον δορυφορικό αιςκθτιρα του Sentinel 5- Precursor , TROPOMI , επιτεφχκθκε για προϊόντα όπωσ το μονοξείδιο του άνκακα και το μεκάνιο εξάγοντασ ζναν ςυντελεςτι ςυςχζτιςθσ

μεταξφ των δορυφορικϊν δεδομζνων και επίγειων μετριςεων με χριςθ φαςματιοςκοπίασ υπερφκρου μζςω μεταςχθματιςμοφ Fourier. Επιπλζον, προςπακιςαμε να ςυγκρίνουμε τα αποτελζςματά μασ μζχρι τϊρα με μετριςεισ από διαφορετικά Ευρωπαικά ι Παγκόςμια δίκτυα.

1.) Introduction

Human nature and environment are intertwined. The need for accurate measures of greenhouse gas emissions is more imperative than ever. In addition to scientific research, we are now pointing out the issue of our survival in this fragile planet. More and more efforts are being done from governments to reduce the concentrations of these dangerous gases to logical levels. The sources of CO2 and CH4 are natural as well as anthropogenic. Due to the rational use of electricity and fossil fuels the concentrations have been increased with an extended rate in the past few years. So , it is very important that we should become environmentally aware of this critical topic.

At first , we are going to start with the impact of these greenhouse gases to climate change and how they contribute to radiative forcing and in which rate. Consequently we will understand the importance of FTIR measurements as it measures greenhouse gases such as carbon monoxide – dioxide- and methane which have great impact on radiative forcing.

The portable EM27/SUN spectrometer is dedicated to measurements of column-averaged abundances of carbon dioxide and methane with sufficient quality for climate research.

1.1) Climate change

Climate change in the world can be caused by various activities. When climate change occurs temperatures can increase dramatically. When temperature rises, many different changes can occur on Earth. For example, it can result in more floods, droughts, or intense rain, as well as more frequent and severe heat waves. Oceans and glaciers have also experienced some changes: oceans are warming and becoming more acidic, glaciers are melting, and sea levels are rising. As these changes

16 frequently occur in future decades, they will likely present challenges to our society and environment.

During the past century, human activities have released large amounts of carbon dioxide and other greenhouse gases into the atmosphere. Most of the gases come from burning fossil fuels to produce energy. Greenhouse gases are like a blanket around the Earth, trapping energy in the atmosphere and causing it to warm. This is called the greenhouse effect and it is natural and necessary to support life on earth. However, while greenhouse gases build-up, the climate changes and result in dangerous effects to human health and ecosystems. A warmer climate can bring changes that can affect our water supplies, agriculture, power and transportation systems, the natural environment, and even our own health and safety. There are some climate changes that are unavoidable and nothing can be done about it. For example, carbon dioxide can stay in the atmosphere for nearly a century, so Earth will continue to warm in the future. (IPCC report , 2007)

Global warming has really taken effect in the world over the last century. The gases that have an influence on the atmosphere are water vapour, carbon dioxide, di-nitrogen-oxide, and methane. Almost 30 percent of incoming sunlight is reflected back into space by bright surfaces like clouds and ice. In the other 70 percent, most is absorbed by the land and ocean, and the rest is absorbed by the atmosphere. The absorbed solar energy heats our planet. This absorption and radiation of heat by the atmosphere is beneficial for life on Earth. Today, the atmosphere contains more greenhouse gas molecules, so more of the infrared energy emitted by the surface ends up being absorbed by the atmosphere. By increasing the concentration of greenhouse gases, we are making Earth’s atmosphere a more efficient greenhouse. Climate has cooled and warmed throughout the Earth history for various reasons. Rapid warming like we see today is unusual in the history of our planet. Some of the factors that have an effect on climate, like volcanic eruptions and changes in the amount of solar energy, are natural. Climate can change if there is a change in the amount of solar energy that gets to the Earth. Volcano eruptions can really affect climate, because when it erupts it spews out more than just lava and ash. Volcanoes release tiny particles made of sulphur dioxide into the atmosphere. These particles get into the stratosphere and reflect solar radiation back out to space. Snow and ice also have a great effect on climate. When snow and ice melts Earth’s climate warms, less energy is

17 reflected and this causes even more warming. There are many different ways that plants, animals, and other life on our planet can affect climate. Some can produce greenhouse gases that trap heat and aid global warming through the greenhouse effect. Carbon dioxide is taken out of the atmosphere by plants as they make their food by photosynthesis. During the night, plants release some carbon dioxide back into the atmosphere. Methane is made while farm animals, such as cattle and sheep digest their food. Cars and trucks can affect climate by releasing carbon dioxide when fossil fuels are burned to power them. When wildfires occur, carbon dioxide is released into the atmosphere. However, if a forest of similar size grows again, about the same amount of carbon that was added to the atmosphere during the fire will be removed. Some effects that scientists have predicted in the past would result when global change was occurring: loss of sea ice, accelerated sea level rise, and more intense heat waves. Scientists have confidence that global temperatures will continue to rise for decades to come, largely due to greenhouse gases produced by human activities. The Intergovernmental Panel on Climate Change (IPCC) stated that the extent climate change effects on individual regions will vary over time and with ability of different societal and environmental systems mitigate or adapt to change (The Intergovernmental Panel on Climate Change). This has been the warmest decade since 1880. According to the National Oceanic and Atmospheric Administration, 2010 and 2005 has been the warmest years on record. The earth could warm by an additional 7.2 degrees Fahrenheit during the 21st century if we fail to reduce emissions from burning fossil fuels (The National Oceanic and Atmospheric Administration).

The rising of temperature will have great effects on the earth’s climate patterns and on all living things. Industrial activities that our modern civilization depends upon have raised atmospheric carbon dioxide from 280 parts per million to 379 parts per million in the last 150 years (The Intergovernmental Panel on Climate Change).

In conclusion, we need to take part and try to stop global warming and other effects on climate change. If the earth’s temperatures continue to rise in the future, living things on earth would become extinct due to the high temperatures. If humans contribute to control global warming, this world would be cooler and the high temperatures we currently have would decrease. If everybody as one take stand and try to end most of

18 the climate changes that are occurring, this world would be a safer place to live on.

Figure 1.1 : Worldwide emissions of carbon dioxide, methane, nitrous oxide, and several fluorinated gases from 1990 to 2010. For consistency, emissions are expressed in million metric tons of carbon dioxide equivalents. These totals include emissions and sinks due to land-use change and forestry. * HFCs are hydrofluorocarbons, PFCs are perfluorocarbons, and SF6 is sulfur hexafluoride. (source John Houghton - Global Warming~ The Complete Briefing – 0521709164)

Figure 1.2 : The amount of radiative forcing caused by various greenhouse gases, based on the change in concentration of these gases in the Earth’s atmosphere. Radiative forcing is calculated in watts per square meter, which represents the size of the energy imbalance in the atmosphere. Data source: NOAA, 2016(source John Houghton - Global Warming~ The Complete Briefing – 0521709164)

Table 1.1.1 : The composition of the atmosphere , the main constituents nitrogen and oxygen and the greenhouse gases

Source Cambridge University Press --- The Edinburgh Building, Cambridge CB2 8RU, UK © J. T. Houghton 1994, 1997, 2004, 2009

1.2) The greenhouse effect

The gases nitrogen and oxygen that make up the bulk of the atmosphere ( Table 1 gives details of the atmosphere’s composition) neither absorb nor emit thermal radiation. It is the water vapour, carbon dioxide and some other minor gases present in the atmosphere in much smaller quantities ( Table 1) that absorb some of the thermal radiation leaving the surface, acting as a partial blanket for this radiation and causing the difference of 20 to 30 °C between the actual average surface temperature on the Earth of about 15 °C and the temperature that would apply if greenhouse gases were absent.

This blanketing is known as the natural greenhouse effect and the gases are known as greenhouse gases. It is called ‘natural’ because all the atmospheric gases (apart from the chlorofluorocarbons – CFCs) were there long before human beings came on the scene. Later on we will

21 mention the enhanced greenhouse effect: the added effect caused by the gases present in the atmosphere due to human activities such as deforestation and the burning of fossil fuels. ( https://climate.nasa.gov/faq/19/what-is-the-greenhouse-effect/)

The basic science of the greenhouse effect has been known since early in the nineteenth century when the similarity between the radiative properties of the Earth’s atmosphere and of the glass in a greenhouse was first pointed out – hence the name ‘greenhouse effect’. In a greenhouse, visible radiation from the Sun passes almost unimpeded through the glass and is absorbed by the plants and the soil inside. The thermal radiation that is emitted by the plants and soil is, however, absorbed by the glass that re-emits some of it back into the greenhouse. The glass thus acts as a ‘radiation blanket’ helping to keep the greenhouse warm. (IPCC report , 2007)

However, the transfer of radiation is only one of the ways heat is moved around in a greenhouse. A more important means of heat transfer is convection, in which less dense warm air moves upwards and more dense cold air moves downwards. A familiar example of this process is the use of convective electric heaters in the home, which heat a room by stimulating convection in it. The situation in the greenhouse is therefore more complicated than would be the case if radiation were the only process of heat transfer. Mixing and convection are also present in the atmosphere, although on a much larger scale, and in order to achieve a proper understanding of the greenhouse effect, convective heat transfer processes in the atmosphere must be taken into account as well as radiative ones. (ref. IPCC (2007))

Figure 1.3 : Schematic of the natural Greenhouse effect (source John Houghton - Global Warming~ The Complete Briefing – 0521709164)

Figure 1.4 : Ice , ocean , land surfaces and clouds all play an important role in determining how much incoming solar radiation the Earth reflects back into space (source John Houghton - Global Warming~ The Complete Briefing – 0521709164)

1.3) The anthropogenic Green-House Effect In addition to the natural greenhouse effect, the rise of atmospheric GHG concentrations for the last 250 years causes an extra atmospheric warming, called the anthropogenic or enhanced greenhouse effect. Since the industrial revolution, CO2 concentrations raised from 260- 280 ppm in 1750 to 379 ppm in 2005 [IPCC, 2007]. The CO2 growth rate will be doubled within the next 50 years, referred to pre-industrial concentrations [IPCC, 2001]. Assuming this doubling of CO2 while concentrations of all other gases remain constant, the long wave radiation reaching space will be reduced by 4.1 W/m2 [Schmidt et al., 2010]. To compensate this, the atmosphere gets warmer with a temperature rise of about 1.2 K ( 10%) in the troposphere and Earth's surface [IPCC, 2001]. Positive feedbacks, like the increase of tropospheric water vapour, cause an additional warming resulting in an overall effect of 1.5 K [IPCC, 2001], which is still under investigation. Uncertainties are due to deficient knowledge about the interaction between radiation and clouds. Supposing an already saturated CO2 absorption, another increase in atmospheric CO2 would not have any additional impacts on climate. But this is not the case. While the core of the 5W/m^2 CO2 band is saturated, the edges of the band remain unsaturated, being the reason for the un-proportional but logarithmic behaviour of radiative forcing related to an increase of CO2. Each CO2 doubling leads to reduced long wave radiation of 4.1 W/m2, where other GHGs have similar effects causing an additional enhancement of this scenario. (source: John Houghton - Global Warming~ The Complete Briefing – 0521709164)

The greenhouse gases are those gases in the atmosphere which, by absorbing thermal radiation emitted by the Earth’s surface, have a blanketing effect upon it. The most important of the greenhouse gases is water vapour, but its amount in the atmosphere is not changing directly because of human activities. The important greenhouse gases that are directly influenced by human activities are carbon dioxide, methane, nitrous oxide, the chlorofluorocarbons (CFCs) and ozone. (source: John Houghton - Global Warming~ The Complete Briefing – 0521709164)

Figure 1.5 : Regions of the infrared spectrum where the greenhouse gases absorb (An Introduction to Atmospheric Radiation – Liu)

The above figure illustrated the regions of the infrared spectrum where the greenhouse gases absorb. Their importance as greenhouse gases depends both on their concentration in the atmosphere ( Table1 ) and on the strength of their absorption of infrared radiation. Both these quantities differ greatly for various gases.

Carbon dioxide is the most important of the greenhouse gases that are increasing in atmospheric concentration because of human activities. If, for the moment ,we ignore the effects of the CFCs and of changes in ozone, which vary considerably over the globe and which are therefore more difficult to quantify, the increase in carbon dioxide (CO2 ) has contributed about 72% of the enhanced greenhouse effect to date, methane (CH4 ) about 21% and nitrous oxide (N2O ) about 7%. (IPPC report), (source: John Houghton - Global Warming~ The Complete Briefing – 0521709164)

Measures imposing substantial reductions in CO2 and other GHG emissions would severely limit global energy use. If GHG emissions are not limited, there are strong scientific grounds to believe they will cause significant global warming in the next few decades. Such a warming would be global, but with regional impacts of varying severity. Concomitant changes are expected to occur in regional patterns of temperature, precipitation and sea level, with resultant impacts on most

25 social, economic and environmental aspects of existence on this planet. With increasing world population, impacts can be expected not only in the areas of agriculture and water resources, affecting food security, but also via the sea-level rise affecting coastal settlements including a number of the world’s major cities. These impacts are expected to impinge on the economy and the welfare of every nation.

1.4) Changes in radiative forcing

Clear evidence has been presented in the previous section that the atmospheric concentrations of a number of radiatively active GHGs have increased over the past 100 years as a result of human activity. By IR absorption, they increase the heat trapping ability of the atmosphere, driving climate change. As discussed above, human-related activities have also been responsible for increases in greenhouse gases in the atmosphere, which can have either heating or cooling effects. This change in the planetary radiation budget is termed the radiative forcing of the Earth’s climate system. Changes in the Earth’s surface temperature are approximately linearly proportional to the radiative forcing inducing those changes, although there is some non-linearity induced by the sensitivity of climate response to height, latitude and the nature of the forcing (e.g.,Jain et al., 2000).

Recent assessments of the direct radiative forcing due to the changes in GHG concentrations are generally in good agreement, determining an increase in radiative forcing of about 2.64 W m^(−2) from the late 1700s to the present time (IPCC, 2007). Of the 2.64 W m^(−2) changes in radiative forcing from greenhouse gases over the past 200 years, approximately 0.7 W m^(−2) occurred within the past decade. By far the largest effect on radiative forcing has been the increasing concentration of carbon dioxide, accounting for about 63% (1.66 W m^−2) of the total forcing. The radiative forcing, due to changes in concentrations of CH4,

N2O, and CFCs and other halocarbons since the late 1700s are 0.48, 0.16 and 0.34 W m^−2, respectively.( IPCC (2007))

Figure 1.6 : IPCC (2007) estimated change in globally and annually averaged anthropogenic radiative forcing (RF; W/m−2) resulting from a number of agents since pre-industrial time (1750) to present (about 2005). LOSU, level of scientific Understanding. (IPCC 2007 REPORT) (source: John Houghton - Global Warming~ The Complete Briefing – 0521709164)

2) Greenhouse gases

Greenhouse gases trap heat and make the planet warmer. Human activities are responsible for almost all of the increase in greenhouse gases in the atmosphere over the last 150 years. The largest source of

27 greenhouse gas emissions from human activities is from burning fossil fuels for electricity, heat, and transportation.

Atmospheric concentrations of both the natural and man-made gases have been rising over the last few centuries due to the industrial revolution. As the global population has increased and our reliance on fossil fuels (such as coal, oil and natural gas) has been firmly solidified, so emissions of these gases have risen. While gases such as carbon dioxide occur naturally in the atmosphere, through our interference with the carbon cycle (through burning forest lands, or mining and burning coal), we artificially move carbon from solid storage to its gaseous state, thereby increasing atmospheric concentrations. (Source : EPA-epa.gov)

2.1) Carbon dioxide

Carbon dioxide enters the atmosphere through burning fossil fuels (coal, natural gas, and oil), solid waste, trees and other biological materials, and also as a result of certain chemical reactions (e.g., manufacture of cement). Carbon dioxide is removed from the atmosphere (or "sequestered") when it is absorbed by plants as part of the biological carbon cycle. (epa.gov).

Carbon dioxide (CO2) is the primary greenhouse gas emitted through human activities. In 2017, CO2 accounted for about 81.6 percent of all greenhouse gas emissions from human activities, (epa.gov). Carbon dioxide is naturally present in the atmosphere as part of the Earth's carbon cycle (the natural circulation of carbon among the atmosphere, oceans, soil, plants, and animals). Human activities are altering the carbon cycle–both by adding more CO2 to the atmosphere, by influencing the ability of natural sinks, like forests, to remove CO2 from the atmosphere, and by influencing the ability of soils to store carbon.

While CO2 emissions come from a variety of natural sources , human- related emissions are responsible for the increase that has occurred in the atmosphere since the industrial revolution. The main sources of CO2 emissions are described below.

 Transportation. The combustion of fossil fuels such as gasoline and diesel to transport people and goods was the largest source of

CO2 emissions in 2017, e.g. accounting for about 34.2 percent of

total U.S. CO2 emissions and 27.7 percent of total U.S. greenhouse gas emissions. This category includes transportation sources such as highway vehicles, air travel, marine transportation, and rail.

 Electricity. Electricity is a significant source of energy and is used to power homes, business, and industry. In 2017 e.g. the combustion of fossil fuels to generate electricity was the second

largest source of CO2 emissions accounting for about 32.9 percent

of total U.S. CO2 emissions and 26.7 percent of total U.S. greenhouse gas emissions. The type of fossil fuel used to generate

electricity will emit different amounts of CO2. To produce a given

amount of electricity, burning coal will produce more CO2 than oil or natural gas.

 Industry. Many industrial processes emit CO2 through fossil fuel

consumption. Several processes also produce CO2 emissions through chemical reactions that do not involve combustion; for example, the production and consumption of mineral products such as cement, the production of metals such as iron and steel, and the production of chemicals. Note that many industrial processes also use electricity and therefore indirectly cause the emissions from the electricity production.

Carbon dioxide is constantly being exchanged among the atmosphere, ocean, and land surface as it is both produced and absorbed by many microorganisms, plants, and animals. However, emissions and removal of CO2 by these natural processes tend to balance, absent anthropogenic impacts. Since the Industrial Revolution began around 1750, human activities have contributed substantially to climate change by adding

CO2 and other heat-trapping gases to the atmosphere.

Source : EPA-epa.gov

Figure 2.1 : Sources of carbon dioxide CO2 ( source http://www.climate-change knowledge.org/ghg_sources.html)

2.2) Methane (CH4)

Methane is emitted during the production and transport of coal, natural gas, and oil. Methane emissions also result from livestock and other agricultural practices and by the decay of organic waste in municipal solid waste landfills.

Methane is also emitted by natural sources such as natural wetlands. In addition, natural processes in soil and chemical reactions in the atmosphere help remove CH4 from the atmosphere. Methane's lifetime in the atmosphere is much shorter than carbon dioxide (CO2), but CH4 is more efficient at trapping radiation than CO2. Pound for pound, the comparative impact of CH4 is more than 25 times greater than CO2 over a 100-year period.

Globally, 50-65 percent of total CH4 emissions come from human activities. Methane is emitted from energy, industry, agriculture, and waste management activities, described below.

 Energy and Industry: Natural gas and petroleum systems are

the largest source of CH4 emissions. Methane is the primary component of natural gas. Methane is emitted to the atmosphere during the production, processing, storage, transmission, and distribution of natural gas and the production, refinement, transportation, and storage of crude oil. Coal mining is also a

source of CH4 emissions.  Agriculture: Domestic livestock such as cattle, swine, sheep, and

goats produce CH4 as part of their normal digestive process. Also, when animals' manure is stored or managed in lagoons or holding

tanks, CH4 is produced. Because humans raise these animals for food and other products, the emissions are considered human- related. When livestock and manure emissions are combined, the

Agriculture sector is the largest source of CH4 emissions.

Methane is emitted from a number of natural sources as well. Natural wetlands are the largest source, emitting CH4 from bacteria that decompose organic materials in the absence of oxygen. Smaller sources include termites, oceans, sediments, volcanoes, and wildfires. Source : EPA-epa.gov

Direct atmospheric measurement of atmospheric methane has been possible since the late 1970s and its concentration rose from 1.52 ppmv in 1978 by around 1 percent per year to 1990, since when there has been little sustained increase. The current atmospheric concentration is approximately 1.77 ppmv, and there is no scientific consensus on why methane has not risen much since around 1990.(IPCC REPORT 2007)

Figure 2.2 : Sources of Methane (CH4) – (source http://www.climate-change knowledge.org/ghg_sources.html)

2.3) Water vapour (H2O)

Water vapour (H2O) is the most abundant greenhouse gas in the atmosphere, which is why it is addressed here first. However, changes in its concentration is also considered to be a result of climate feedbacks related to the warming of the atmosphere rather than a direct result of industrialization. The feedback loop in which water is involved is critically important to projecting future climate change, but as yet is still fairly poorly measured and understood.

As the temperature of the atmosphere rises, more water is evaporated from ground storage (rivers, oceans, reservoirs, soil). Because the air is

32 warmer, the absolute humidity can be higher (in essence, the air is able to 'hold' more water when it's warmer), leading to more water vapour in the atmosphere. As a greenhouse gas, the higher concentration of water vapour is then able to absorb more thermal IR energy radiated from the Earth, thus further warming the atmosphere. The warmer atmosphere can then hold more water vapour and so on and so on. This is referred to as a 'positive feedback loop'. However, huge scientific uncertainty exists in defining the extent and importance of this feedback loop. As water vapour increases in the atmosphere, more of it will eventually also condense into clouds, which are more able to reflect incoming solar radiation (thus allowing less energy to reach the Earth's surface and heat it up). The future monitoring of atmospheric processes involving water vapour will be critical to fully understand the feedbacks in the climate system leading to global climate change. As yet, though the basics of the hydrological cycle are fairly well understood, we have very little comprehension of the complexity of the feedback loops. Also, while we have good atmospheric measurements of other key greenhouse gases such as carbon dioxide and methane, we have poor measurements of global water vapour, so it is not certain by how much atmospheric concentrations have risen in recent decades or centuries, though satellite measurements, combined with balloon data and some in-situ ground measurements indicate generally positive trends in global water vapour. Water evaporation from the ocean produces about 67% of Earth’s water vapour, or approximately 505,000 km3 (121,000 mi^3 ) of water, about 398,000 km3 (95,000 mi^3 ) of it over the oceans(IPCC REPORT 2013). On-going evaporation helps cool the ocean (as heat is removed from the ocean’s surface water molecules and transformed into gaseous water molecules). Without this natural process, the ocean would heat up and, in turn, cause global air temperatures to rise.

Source : EPA-epa.gov

Figure 2.3 : Ocean waves (source http://www.ox.ac.uk/news/2015-12-16-freak- ocean-waves-hit-without-warning-new-research-shows)

2.4) Carbon monoxide (CO)

Carbon monoxide (CO) is only a very weak direct greenhouse gas, but has important indirect effects on global warming. Carbon monoxide reacts with hydroxyl (OH) radicals in the atmosphere, reducing their abundance. As OH radicals help to reduce the lifetimes of strong greenhouse gases, like methane, carbon monoxide indirectly increases the global warming potential of these gases.

Carbon monoxide in the atmosphere can also lead to the formation of the tropospheric greenhouse gas 'ozone'. Atmospheric concentrations of carbon monoxide vary widely around the world and throughout the year, ranging from as low as 30 parts per billion up to around 200 parts per billion. Concentrations increased during the 20th century, but there are some signs that concentrations dropped slightly in the 1990s due to widespread use of catalytic convertors, with their lower carbon monoxide emissions, in cars.

Aside from man-made sources, a great deal of carbon monoxide comes from the chemical oxidation of methane and other hydrocarbons in our atmosphere. Additional natural sources include emission from vegetation and the world's oceans. By far the largest sink for carbon monoxide is its reaction with OH in the atmosphere. However, a small but significant amount is also lost from the atmosphere through deposition on the ground.( Source : EPA-epa.gov)

2.4.1) Human Impact Today more than half of carbon monoxide emissions are man-made. The highest concentrations of carbon monoxide tend to occur close to areas of high human population. On a global scale, this has meant that the more densely populated northern hemisphere has higher concentrations of carbon monoxide than the southern hemisphere. Biomass burning

35 and fossil fuel use are the main sources of man-made carbon monoxide emissions.

Breathing air with a high concentration of CO reduces the amount of oxygen that can be transported in the blood stream to critical organs like the heart and brain. At very high levels, which are possible indoors or in other enclosed environments, CO can cause dizziness, confusion, unconsciousness and death. Very high levels of CO are not likely to occur outdoors. However, when CO levels are elevated outdoors, they can be of particular concern for people with some types of heart disease. These people already have a reduced ability for getting oxygenated blood to their hearts in situations where the heart needs more oxygen than usual. They are especially vulnerable to the effects of CO when exercising or under increased stress. In these situations, short-term exposure to elevated CO may result in reduced oxygen to the heart accompanied by chest pain also known as angina.( Source : EPA-epa.gov)

2.4.2) Potential for control As with many direct and indirect greenhouse gases, reductions in carbon monoxide emissions can most effectively be made through direct reductions in fossil fuel use. There is some evidence that the widespread use of catalytic convertors in cars has significantly reduced carbon monoxide emissions from this source. However, such reductions must be balanced against the increased emissions of the greenhouse gases carbon dioxide and nitrous oxide which often result from a switch to catalytic convertors.

Figure 2.4 : Sources of Carbon Monoxide (CO) Emission (Source: U.S. EPA-epa.gov)

Source : EPA-epa.gov

3.1) Interaction of radiation and matter

Energy is absorbed or emitted by a molecule, thereby changing its electronic, vibrational and rotational state. As the energy needed to excite a molecule is highly species-specific, from the absorption characteristics information about the structure of the molecule or physical processes can be gained. Molecular spectroscopy is therefore an important technique in many fields of science, for example physics, chemistry, medical science, biology, astronomy and environmental studies. For our case this technique is used to derive concentrations of trace gases in the atmosphere.

3.1.1) Molecular Spectroscopy

Molecular spectroscopy is the discipline which investigates the interactions of molecules with electromagnetic radiation. Absorption, emission, or scattering of electromagnetic radiation by molecules or molecular ions can be used to study molecular structures or physical processes in which molecules are involved. Analyzing molecules using spectroscopic method sis widely used in many scientific fields including physics, chemistry, biology, astronomy and environmental sciences. In the latter, it is an established technique for detecting and measuring the concentrations of molecules in the Earth’s atmosphere. In general, molecular spectral lines result from internal changes of energetic states of molecules, namely electronic, translational, vibrational and rotational transitions. The resulting spectral signatures are characteristic for each kind of molecule and provide an excellent scientific tool to detect and identify molecules. In the next sections, the formation of spectral lines is described for diatomic molecules. In analogy to atoms, the same quantum mechanical principles hold for molecules. However, molecular spectra are more complex due to additional internal degrees of freedom which results in additional quantum numbers.

3.1.2) Molecular vibrations

Atoms within a molecule are constrained by molecular bonds to move together in a certain specified ways, called degrees of freedom that can be: electronic, translational, rotational and vibrational. In electronic motion, the electrons change energy levels or directions of spins. The translational motion is characterized by a shift of an entire molecule to a new position. The rotational motion is described as a rotation of the molecule around its centre of mass. When the individual atoms within a molecule change their relative position then we say that the molecule vibrates. If we have a nonlinear molecule consisting of N atoms, we need to specify 3N coordinates that correspond to their locations. Three of those can be used to specify the centre of mass of the molecule, leaving 3N-3 coordinates for the location of the atoms relative to the centre of mass. For determining the orientation of the molecule we need to specify three angles (if the molecule is linear, only two angles are sufficient), so

38 leaving 3N-6 coordinates that, when varied, do not change the location of the centre of the mass nor the orientation of the molecule. These 3N- 6 coordinates correspond to different vibrational degrees of freedom of the molecule that can range from the simple coupled motion of the two atoms of a diatomic molecule to the much more complex motion of each atom in a large poly-functional molecule. When a molecule is exposed to wide-spectrum radiation, some distinct parts of it are absorbed by the molecule. The absorbed wavelengths are the ones that match the transitions between the different energy levels of the corresponding degrees of freedom of that molecule. The vibrational transitions are the most important transitions for IR spectroscopy because IR radiation is too low to affect the electrons within the individual atoms and too powerful for rotational and translational transitions. Source (Source : FOURIER TRANSFORM INFRARED SPECTROSCOPY Avtor: Mimoza Naseska Mentor: assoc. prof. dr. Matjaz Zitnik Ljubljana, March 2016)

3.1.3) Molecular Potential

In the case of vibrational transitions, the absorption of the radiation by the molecule can be described in terms of a resonance condition. The specific oscillating frequency of the absorbed radiation matches the natural frequency of a particular normal mode of the molecular vibration. The simplest model for describing the molecular vibrations is the one of a diatomic molecule where the bond of the two atoms is approximated by a weightless spring. The force needed to move the atoms by a certain distance x from an equilibrium position is proportional to the force constant k, which measures the strength of the bond. That is the Hooke's law:

F = -kx

According to the Newton's law the force is also proportional to the mass m and its acceleration, the second derivative of the distance x with respect to time t:

If we combine the two equations above, we get a second order differential equation:

with a solution:

that describes the motion of the atoms as a harmonic oscillation. In the equation above v is the vibrational frequency and φ is the phase angle. In the case of diatomic molecule, the frequency of vibration is given with the equation:

Where μ is the reduced mass of a diatomic molecule defined as: 1/μ = (1/m1) + (1/m2)m1 and m2 are the masses of the individual atoms making up the molecule.

 The higher the force constant K  the higher the frequency v !

Figure 3.1: Absorbance vs wavenumber  Source : BrukerOptics

 The higher the vibrating atomic mass  the lower the vibrational frequency v!

Figure 3.2: Absorbance vs wavenumber with higher the vibrating atomic mass  Source : BrukerOptics

The potential energy of a molecule obeying the Hooke's law is obtained by integrating the equation :

F = -kx, because F = -(dV/dx ) :

Figure 3.3 : Mechanical model of a vibrating diatomic molecule- Source : BrukerOptics

The graph of the function below, is a parabola, it is referred to as a harmonic potential because the molecule performs a harmonic motion. According to classical mechanics, a harmonic oscillator may vibrate with any amplitude, which means that it can possess any amount of energy until it is confined in the potential well. Quantum mechanics shows that molecules can only exist in definite energy states. In the case of harmonic potentials the state's energies are equidistant:

Figure 3.4 : The potential energy diagram comparison of the an-harmonic and the harmonic oscillator. Vibration energy states are denoted by n. Source : BrukerOptics.com

Figure 3.5 : The harmonic potential which we have used as a model for describing the atomic motion in a molecule is only a small distance approximation for the real molecular potential. The real potential between atoms in a molecule is an- harmonic i.e. the distance between energy levels decreases with increasing energy while the energy levels of the particle in the harmonic potential are equidistant.

(Source : FOURIER TRANSFORM INFRARED SPECTROSCOPY Avtor: Mimoza Naseska Mentor: assoc. prof. dr. Matjaz Zitnik Ljubljana, March 2016)

There are many types of electromagnetic radiation in the universe besides the mid-infrared, the collection of which is called the electromagnetic spectrum. A diagram of part of the electromagnetic spectrum is seen in next figure. The mid-infrared has been intentionally placed at the center of figure. To the right of the mid-infrared between 400 and 4 cm^(−1) is the far infrared. When molecules absorb far infrared light they vibrate. Molecules with heavy atoms in them, including many inorganics, absorb

43 in this region. Some FTIRs work in the far infrared. Lower in energy than the far infrared are microwaves. When molecules absorb microwaves they rotate faster. Microwave spectra of rotating gas phase molecules have been measured, and this type of spectroscopy can be used to identify and quantitate gases in samples. A microwave oven gives off radiation tuned to an absorbance of liquid water. The liquid water molecules in food absorb this energy and rotate rapidly. Collisions with neighbouring food molecules transfer energy to them raising their temperature and making your dinner warm. Higher in energy than the mid-infrared, from 14,000 to 4000 cm^(−1), is the near infrared. Molecules vibrate when they absorb near infrared radiation, but the spectral features are fewer, broader, and more difficult to interpret than in the mid-infrared.

Because of certain instrumental advantages, near infrared radiation is frequently used to measure sample properties in difficult environments such as in the middle of a chemical reactor or of liquid flowing through a pipe. Some FTIRs work in the near infrared. Higher in energy than the near infrared are visible light and ultraviolet (UV) radiation. These higher energy light waves fall above 14,000 cm^(−1). When a molecule absorbs visible or ultraviolet light the electrons in the molecule transition from a lower electronic energy level to a higher one. Many molecules have measurable UV and visible spectra, and these spectra can be used to identify and quantitate molecules in samples. FTIRs can be equipped to work in the visible and UV. Figure 3.6 shows that as you move from right to left across the electromagnetic spectrum there is an increase in energy, wavenumber, and frequency, but a decrease in wavelength. Similarly, moving from left to right across Figure 3.6 there is a decrease in energy ,frequency, and wavenumber and an increase in wavelength.

Source : 2011 by Taylor and Francis Group, LLC ,CRC Press is an imprint of Taylor & Francis Group, an Informa business

Figure 3.6 : A part of electromagnetic spectrum Source : 2011 by Taylor and Francis Group, LLC ,CRC Press is an imprint of Taylor & Francis Group, an Informa business

3.2 ) Discovery of Infrared light

In 1800 the astronomer Friedrich Wilhelm Herschel analyzed the spectrum of sunlight. Herschel created the spectrum by directing sunlight through a glass prism so that the light was divided into its different colours. He measured the heating ability of each color using thermometers with blackened bulbs. When he measured the temperature just beyond the red part of the spectrum he noticed some kind of invisible radiation. Much to his surprise he found that the area close to the red part had the highest heating ability of all. Herschel concluded that there must be a different kind of light beyond the red portion of the spectrum , which is not visible to human eye. This kind of light became known as “ Infrared “ ( below red ) light.

Herschel then placed a water-filled container between the prism and thermometer and observed that the temperature measured was slower

45 than the one measured without the water. Consequently , the water must partially absorb the radiation.

In addition , Herschel could prove that depending on how the prism was rotated ( i.e. depending on the spectral range ) the difference in the temperatures measured for each color varied. This was the beginning of Infrared spectroscopy!

Figure 3.7 : Electromagnetic spectrum and “Herschel’s” band  Source (http://herschel.cf.ac.uk/science/infrared)

Recent developments in detector technology have led to many useful applications using infrared radiation. Medical infrared technology is used for the non-invasive analysis of body tissues and fluids. Infrared cameras are used in police and security work, as well as in military surveillance. In fire fighting, infrared cameras are used to locate people and animals caught in heavy smoke, and for detecting hot spots in forest fires. Infrared imaging is used to detect heat loss in buildings, to test for stress and faults in mechanical and electrical systems, and to monitor pollution. Infrared satellites are routinely used to measure ocean temperatures, providing an early warning for El Niño events that usually impact climates worldwide. These satellites also monitor convection within clouds, helping to identify potentially destructive storms. Airborne and space-based cameras also use infrared light to study

46 vegetation patterns and to study the distribution of rocks, minerals and soil. New and fascinating discoveries are being made about our universe in the field of infrared astronomy. The universe contains vast amounts of dust, and one way to peer into the obscured cocoons of star formation and into the hearts of dusty galaxies is with the penetrating eyes of short-wavelength infrared telescopes. Our Universe is also expanding as a result of the Big Bang, and the visible light emitted by very distant objects has been red-shifted into the infrared portion of the electromagnetic spectrum.

Figure 3.8 : Friedrich W. Herschel’s experiment (Source- www.Brukeroptics.com)

3.3 ) What is an Infrared spectrum ?

A plot of measured infrared light intensity versus a property of light is called an infrared spectrum. An example of an infrared spectrum is shown in Figure 3.9. By convention the x-axis of an infrared spectrum is plotted with high wavenumber to the left and low wavenumber to the right. Plots of your FTIR spectra should always follow this convention. Note in Figure 3.9 that 4000 cm^(−1) is to the left and 500 cm^(−1) is to the right, and that the spectrum is plotted in Absorbance units, which

47 measure the amount of light absorbed by a sample. As you can see in the figure the peaks point up and their tops denote wavenumbers at which significant amounts of light were absorbed by the sample. The absorbance spectrum of a sample is calculated from the following equation:

Where ,

A = Absorbance I0 = Intensity in the background spectrum I = Intensity in the sample spectrum

Absorbance is also related to the concentration of molecules in a sample via an equation called Beer’s Law: A = ε*l *c

Where,

A = Absorbance ε = Absorptivity l = Path-length c = Concentration

The height or area of a peak in an absorbance spectrum is proportional to concentration, which is why Beer’s Law can be used to determine the concentrations of molecules in samples.

The y-axis of an infrared spectrum can also be plotted in units called percent transmittance (%T), which measures the percentage of light transmitted by a sample %T spectra are calculated as follows: %T = 100 × (I/Io)

In the next figure we can see how a transmittance spectrum looks like.

Figure 3.9 :The infrared spectrum of polystyrene plotted in percent transmittance (%T) ( source http://www.ifsc.usp.br )

3.4 ) What kind of molecules absorb infrared light ?

Infrared light can only be absorbed by a molecule if the dipole moment of the specific group of atoms changes during the vibration. This absorption of energy resulting in an increase in the amplitude of the vibrations is known as resonance. The greater the change in dipole moment, the strongest the corresponding IR absorption band will be!

Source : BrukerOptics

Vibrations not accompanied by changes in the dipole moment can not be excited by absorption of IR light and are termed IR inactive. A consequence of this is that homo-nuclear diatomic molecules don’t have any IR spectrum.

Source : BrukerOptics

4) The Fourier Transform Principal

Fourier transform is a mathematical technique that can be used to transform a function from one real variable to another. It is a unique powerful tool for spectroscopists because a variety of spectroscopic studies are dealing with electromagnetic waves covering a wide range of frequency. In Fourier transform an important role plays the term , when x represents frequency and the corresponding y is time. This provides an alternate way to process signal in time domain instead of the conventional frequency domain. To realize this idea, Fourier transform from time domain to frequency domain is the essential process that enable us to translate raw data to readable spectra. Recent prosperity of Fourier transform in spectroscopy should also attribute to the development of efficient Fast Fourier Transform algorithm.

Source(https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Te xtbook_Maps/Supplemental_Modules_(Physical_and_Theoretical_Chemistry)/Spectroscopy /Vibrational_Spectroscopy/Infrared_Spectroscopy/How_an_FTIR_Spectrometer_Operates )

. The Fourier Transform is used EVERYWHERE!!!

 It is being used right know on your cell phones to filter out noise from your voice

 It is used to compress digital photos

 It is used to analyze financial data

 It is used to interpret radar signals

. The Fourier Transform has been called the most important math algorithm

The first one who found that a spectrum and its interferogram are related via a Fourier transform was Lord Rayleigh. He made the discovery in 1892. But the first one who successfully converted an interferogram to its spectrum was Fellgett who made the accomplishment after more than half a century. Fast Fourier transform method on which the modern FTIR spectrometer based was introduced to the world by Cooley and Turkey in 1965. It has been applied widely to analytical methods such as infrared spectrometry, nuclear magnetic resonance and mass spectrometry due to several prominent advantages which are listed in Table 4.1.1.

Table 4.1.1: Advantages of Fourier Transform over Continuous-Wave Spectrometry (https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Textbook_ Maps/Supplemental_Modules_(Physical_and_Theoretical_Chemistry)/Spectroscopy/Vibrati onal_Spectroscopy/Infrared_Spectroscopy/How_an_FTIR_Spectrometer_Operates )

The principle of Fourier transformation (FT) is used for the analysis of measurement signals. The received intensity of the beam as a function of optical path difference x can be converted into a spectrum. Considering a monochromatic source of light, the two interfered wave trains show a phase difference depending on x. The beams interfere constructive if x is a multiple of the wavelength:x = nλ with n = 0,1,2,3 ... where the detected intensity is equal to the intensity of the source.

Therefore the resulting interferogram I’ (x) is a harmonic oscillation and can be described as a cosine-function:

If the light of the source is not monochromatic but contains many wavenumbers (asthe Sun), the correct intensity is given by integration over all wavenumbers.

 Fourier-transform infrared spectroscopy is a vibrational spectroscopic technique, meaning it takes advantage of asymmetric molecular stretching, vibration, and rotation of chemical bonds as they are exposed to designated wavelengths of light.

 Two widely-spaced lines in spectrum give an interferogram which repeats over a short distance. Taking data over a short path difference (time) is sufficient to resolve the lines.

 Two close lines give an interferogram which repeats over a long distance (because the cosine waves are nearly in phase).

The interferogram must be measured over a longer path difference (time) to get a satisfactory spectrum.

Fourier transform, named after the French mathematician and physicist Jean Baptiste Joseph Fourier, is a mathematical method to transform a function into a new function. The following equation is a common form of the Fourier transform with unitary normalization constants:

F(ω)= ∫ √

52 in which t is time, i is the square root of -1.

The following equation is another form of the Fourier transform (cosine transform) which applies to real, even functions:

∫ √

The following equation shows how f(t) is related to F(v) via a Fourier transform:

∫ √

The interferogram obtained is a plot of the intensity of signal versus OPD. A Fourier transform can be viewed as the inversion of the independent variable of a function. Thus, Fourier transform of the interferogram can be viewed as the inversion of OPD. The unit of OPD is centimeter, so the inversion of OPD has a unit of inverse centimeters, cm-1. Inverse centimeters are also known as wavenumbers. After the Fourier transform, a plot of intensity of signal versus wavenumber is produced. Such a plot is an IR spectrum.

(https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Textbook_ Maps/Supplemental_Modules_(Physical_and_Theoretical_Chemistry)/Spectroscopy/Vibrati onal_Spectroscopy/Infrared_Spectroscopy/How_an_FTIR_Spectrometer_Operates )

Figure 4.1 : An example of an IR Transmission spectrum (in the mid-IR range 100- 4000)  Source : BrukerOptics.com

4.1) EM/27 SUN BRUKER INSTRUMENT

EM27/SUN is a mobile analyzer dedicated for atmospheric studies at high performance. It utilises an innovative camera-based feedback system to follow the movement of the sun which is used as a light source. The high accuracy and the robots tracking is the basis for reliable and high precision gas analysis. It uses the rock solid Interferometer , so the system is very robust and always aligned. (Bruker.com)

Figure 4.2 : EM27/SUN FTIR on the terrace of AUTH in Thessaloniki ( Credit Konstantinos Michailidis )

The EM27/SUN spectrometer is intended to measure direct solar radiation in the near infrared (NIR) spectral range.

The recorded spectra contain signatures of atmospheric constituents (H2O, CO2, CH4, O2), which can be evaluated to retrieve the total columns.

The EM27/SUN contains a solar tracker, which is used to reflect the solar radiation onto the detector. A very high precision of the tracker is reached by using a camera and image processing software to detect the actual line of sight of the spectrometer and to correct the solar tracker angles if necessary.

With 10 scans and a scanner velocity of 10 kHz, one measurement takes about 58 s.

Figure 4.3 : Schematic drawing of the tracker used. (Picture taken from Huster, 1998.)-Source(Camtracker: a new camera controlled high precision solar tracker system for FTIR-spectrometers M. Gisi, F. Hase, S. Dohe, and T. Blumenstock) ( right panel ) Typical Thessaloniki spectrum recorded by the InSb and InGaAs diode. (left panel).

Figure 4.4 : EM27/SUN BRUKER instrument in Thessaloniki

The EM27/SUN spectrometer, which was developed by KIT in collaboration with Bruker-Optics-TM, is utilized for the acquisition of

56 solar spectra. Central part of this Fourier transform spectrometer (FTS) is a RockSolidTM pendulum interferometer with two cube corner mirrors and a CaF2 beamsplitter. The EM27/SUN routinely records double sided interferograms, the compensated BS design minimizes the curvature in the phase spectrum. This setup achieves high stability against thermal influence sand vibrations. The retro-reflectors are gimbal-mounted, which results in frictionless and wear-free movement. In this aspect the EM27/SUN is more stable than the HR125 high resolution FTS, which suffers from wear because of the use of friction bearing son the moving retro-reflector. Over time this leads to shear misalignment and requires regular realignment (Hase, 2012). The gimbal-mounted retro-reflectors move a geometrical distance of 0.45 cm leading to an optical path difference of 1.8 cm which corresponds to a spectral resolution of 0.5 cm^(−1), (Hase et. al 2012).

A solar brightness fluctuation correction is performed. Furthermore, the recorded interferograms are Fourier transformed using the Norton-Beer- Medium apodisation function (Daviset al., 2010). This apodisation is useful for reducing side-lobes around the spectral lines, an undesired feature in low resolution spectra, which would complicate the further analysis. A quality control, which filters interferograms with intensity fluctuations above 10 % and intensities below 10 % of the maximal signal range, is also applied.

In this thesis, spectra were analyzed using a spectral fitting algorithm (prf96-EM27-fast), which gives the user the opportunity to provide the measured ILS as an input parameter, an option chosen for this study (Hase et al., 2004). In our case we choose for the ILS parameters ( apodization function and phase error ) the values 0.9862 and 0.0019 for primary and secondary channel too. This code (prf96-EM27-fast) is in wide use and has been thoroughly tested in the past for the EM27/SUN, e.g. Schneider and Hase (2009); Sepúlveda et al. (2012); Kiel et al. (2016a); Chen et al. (2016).

For the evaluation of the O2 column the 7765 - 8005 cm^(−1) spectral region is used, which is also applied in the TCCON analysis (Wunch et al., 2010). For CO2 we combine the two spectral windows used by TCCON to one larger window ranging from 6173 to 6390 cm−1. CH4 is evaluated in the 5897 - 6145 cm^(−1) spectral domain. For H2O the 8353 - 8463 cm^(−1) region is used.

Finally , we use the spectral region (4208.7,4318.8) for the evaluation of CO column and last but not least the spectral band from Tropomi sensor in Sentinel S5-P.

4.2) Theory of EM27/SUN – BRUKER

The EM27 achieves 1.8 cm optical path difference equivalent to 0.5 cm^(−1) resolution. To realise this, the retro-reflectors of the pendulum structure move a geometric distance of 0.45 cm. The sampling of the interferogram is controlled by a standard, not frequency-stabilised HeNe laser. The sealed spectrometer compartment contains a desiccant cartridge where the radiation enters through a wedged fused silica window. So the spectrometer withstands and can be operated in adverse environmental conditions, e.g., high relative humidity.

The tracker leads the solar radiation into the EM27 spectrometer. We limit the beam to 25mm free diameter by using a 750 nm long-pass-filter a few centimetres behind the entrance window in order to block unwanted radiation which would disturb the laser detectors of the interferometer , (Gisi et al. 2011).

After the radiation has passed the interferometer, it is blocked by an aperture with a diameter of 3 mm. This aperture is being inserted for controlling optical aberrations for avoiding nonlinear detector response.

The remaining part of the radiation is appropriate for recording solar absorption spectra. Finally the beam is focused on a field stop with 0.6mm diameter by a 90 degrees off-axis paraboloid with an effective focal length of 101.6 mm. In combination, this results in a semi Field of- View (FOV) of 2.96 mrad. This is half of the angle subtended by the FOV. The field stop is inclined versus the optical axis by a few degrees to avoid channelling. (Hase et al. 2011)

As the detectors show unwanted nonlinear and ILS effects, a diffuser was placed between the field stop and the InGaAs detector (size 1×1mm2, spectral sensitivity 6000–9000 cm^(−1)). The image of the sun on the field stop is recorded by the camera which then is used as the optical feedback for the tracker (CamTracker) (Gisi et al. 2011). This is a very precise method for controlling the actual line of sight of the spectrometer.

Figure 4.5 : The solar tracker of the EM27/SUN ( Thessaloniki 2019 , credit : Konstantinos Michailidis )

Figure 4.6 : Top-view of the camera set-up and the light path in front of and inside the spectrometer.. Source  (Camtracker: a new camera controlled high precision solar tracker system for FTIR-spectrometers M. Gisi, F. Hase, S. Dohe, and T. Blumenstock)

In general, FTIR spectroscopy is a two beam interferometry on the basis of a Michelson interferometer ( which will be presented more precisely in the following section ), being the heart of the spectrometer. In case of groundbased atmospheric measurements, the Sun is the light source and the absorption during its path through the atmosphere is measured. The Sun's radiation is captured by the solar tracker on top of the container and is reflected by two movable mirrors downwards into the container. After passing three fixed mirrors, the light is focused on the input field stop, ensuring that only light of the solar disc's center is considered. This is controlled by a camera based system. Inside the spectrometer, the radiation reaches the beam-splitter and is divided into two beams.

One of them is reflected by a fixed mirror, while the other one is reflected by a moving mirror. Both beams recombine at the beamsplitter, where the optical path difference x depends on the position of the movable mirror. The beam passes several other optical components and is finally focused on the detector, which determines the intensity as a function of the optical path difference, where the resulting AC part of the signal is called an interferogram.

For an arbitrary large x, the resulting interferogram would contain all the spectral information. Since this is not the case, the information content is limited causing a finite resolution. A single mode helium-neon (HeNe) laser is additionally used to exactly determine the optical path difference. The laser beam has the same light path as the sunlight and each zero crossing of the laser's interferogram represents a sampling point. To improve the signal to noise ratio, two or more scans are combined for each measurement.

Compared to a scanning monochromator, one of the most important advantages ofFTIR spectroscopy is the saved time due to the simultaneous detection of all wavelengths.

Figure 4.7 : Daily spectrum ( February 2019 ) in Thessaloniki -- two channels observed , grey is the co channel

4.3) Utility and advantages • Is mainly used by universities and research groups to measure the human impact of greenhouse gas concentrations in urban areas

• Also to investigate the impact of natural gas sources such as bush fires or gases emitted from melting permafrost or from swamps in remote areas

Figure 4.8 : Greater concentrations of greenhouse gases mean more solar radiation is trapped within the Earth’s atmosphere, making temperatures rise. Source: W. Elder, NPS.https://climatechange.lta.org/get- started/learn/co2-methane-greenhouse-effect/

4.4) What is an FTIR spectrum used for ? • An FTIR spectrum is like a chemical fingerprint

• It can be used to characterize new materials or identify and verify known and unknown samples

• Identification and Quantification of Pollutants and Hazardous Material

• High-Precision Quantification of Trace Gases

• Air Quality Monitoring

• Cloud Imaging

• Emission Measurement

• Especially useful in chemical and manufacturing industries , research and product development

• Depending on the spectrometer setup and technique many analytical questions across manifold industries can be answered by FTIR spectroscopy

FTIR spectrometers have several prominent advantages:

 FTIR spectrometers are the third generation infrared spectrometer and have the following advantages , such as:

i. The signal-to-noise ratio of spectrum is significantly higher than the previous generation infrared spectrometers.

ii. The accuracy of wavenumber is high. The error is within the range of ± 0.01 cm-1.

iii. The scan time of all frequencies is short (approximately 1 s).

iv. The resolution is extremely high (0.1 ~ 0.005 cm-1).

v. The scan range is wide (1000 ~ 10 cm-1).

vi. The interference from stray light is reduced.

Due to these advantages, FTIR Spectrometers have replaced dispersive IR spectrometers.

. In spite of the advantages of FT-IR over dispersive instruments such as: high speed data collection, increased resolution, lower detection limits and greater energy throughput, the acceptance of FT-IR spectroscopy was slowed by the complexity of the calculation required to transform the measured data into a spectrum. With the discovery of the Fast Fourier Transform algorithm by James Cooley and John Tukey (1964), the time for spectrum calculation was reduced from hours to just a few seconds. The development of the rest commercial FT-IR spectrometer in 1969 by Digilab enabled spectro-scopists to see a spectrum plotted shortly after the interferogram was collected.

In the next figures we can see the diference between a dispercive infrared spectrometer and a FTIR spectrometer!

Figure 4.9 : Dispersive IR spectrometer (picture by bruker.com)

Figure 4.10 : FTIR spectrometer (picture by bruker.com)

4.5) Interferometer and Fourier Transform

Introduction

The beating heart of our FTIR spectrometer is Michelson’s Interferometer. Fourier transform infrared spectrometer doesn’t select a wavelength of light shine that on the detector and measure the power of that particular segment of light and therefore determine which wavelengths of light are making it through the spectrum. Actually looks at all the lights simultaneously using a broadband light detector that detects all the wavelengths and just tells you us how much light hitting the detector. But we can change that power by creating interferences within the light by using an interferometer and these interference patterns are going to tell us about every single wavelength that exists simultaneously in that light. We could compare this operation with listening to music. If we wanted to listen to a symphony we could use a special maschine to sort out the sounds and just tell us when middle C plays for example. We would sit there and listen to the entire symphony and we would hear middle C whenever it played and then look every note individually or we could listen to the whole symphony. There are some people that can tell you every note in that particular symphony and can text transcribe. These are the people that have a Fourier Transform Infrared Spectrometer in their brain.

4.5.1) Michelson’s Interferometer

Albert Michelson ( 1852 – 1931 ) :

The German-born US physicist Albert Michelson was the first American to win a Nobel Prize. He was awarded this honor in 1907 for his measurements of the speed of light.

However , he is most famous for his role in Michelson-Morley experiment. It had long been thaught that light travelled through a medium called ether ,which was thaught to occupy all space. In the 1880’s Michelson and the US physicist Edward Morley (1838-1931) tried to use the interference of light waves to find out how fast the Earth moved through the ether. However their results suggested that ether didn’t exist at all. This later provided the basis for Einstein’s theory of relativity.

Figure 4.11 : Albert Abraham Michelson (1852-1931), German-born American physicist and winner of the 1907 Nobel Prize for physics. He invented the Michelson interferometer, made precise measurements of the speed of light, and was the first to measure the angular diameter of a star. In conjunction with Edward Morley, he performed in 1887 the Michelson-Morley experiment to detect the motion of the Earth through the postulated ether (a medium believed to be necessary for the propagation of light). The failure to detect any such motion indicated the non-existence of the ether, and led Einstein to his theory of relativity. Credit : SCIENCE PHOTO LIBRARY

Interferometers generally are used to measure very small displacements by using the wave property of light (or other radiation e.g. low energy neutrons). They measure changes of the interference pattern when waves with different phases overlap. While in a spectrometer the displacements between the sources are known (e.g. in a grating) one can

68 determine wave properties e.g. the wave length. If, on the other hand, the wavelength is known one can use this principle to measure displacements of the order of the wavelength of the light used.

The Michelson interferometer produces interference fringes by splitting a beam of monochromatic light so that one beam strikes a fixed mirror and the other a movable mirror. When the reflected beams are brought back together, an interference pattern results.

But first lets see what is an Interference pattern…

4.5.2) Interference Pattern To better understand how interferometers work, it helps to understand more about 'interference'. If you have ever thrown stones into a flat, glassy pond or pool and watched what happened, you already know about interference. When the stones hit the water, they generate concentric waves that move away from the stone's point of entry. And where two or more of those concentric waves intersect, they interfere with each other, the poitns of intersection being larger or smaller or completely canceling each other out. The visible pattern occurring where waves intersect constitutes an "interference" pattern.

The principles of interference are simple to understand. The figure at right shows two specific kinds of interference: total constructive interference and total destructive interference. In total constructive interference, when the peak of one wave merges with the peak of another wave, they add together and 'construct' a larger wave. In total destructive interference, the peak of one wave meets the valley of an identical wave, and they totally cancel each other out (they 'destroy' each other).

In nature, the peaks and valleys of one wave will not always perfectly meet the peaks or valleys of another wave like the illustration shows. Regardless of how they merge, the height of the wave resulting from the interference always equals the sum of the heights of the merging waves. When the waves don't meet up perfectly, partial constructive or destructive interference occurs. The animation below illustrates this effect.

Note how it continues to change as long as the waves continue to interact.

Figure 4.12 : Interference patterns in water. The "interference" occurs in the regions where the expanding circular waves from the different sources intersect. (Credit: Wikimedia commons)

Figure 4.13 : When the peaks of two waves meet, their peaks add up. When the peaks of one wave meet the valleys of another, they cancel out.

Credit: www.explainthatstuff.com

4.5.3) Michelson’s Interferometer principle The Michelson interferometer, which is the core of FTIR spectrometers, is used to split one beam of light into two so that the paths of the two beams are different. Then the Michelson interferometer recombines the two beams and conducts them into the detector where the difference of the intensity of these two beams are measured as a function of the difference of the paths.

A typical Michelson interferometer consists of two perpendicular mirrors and a beamsplitter. One of the mirror is a stationary mirror and another one is a movable mirror. The beamsplitter is designed to transmit half of the light and reflect half of the light. Subsequently, the transmitted light and the reflected light strike the stationary mirror and the movable

71 mirror, respectively. When reflected back by the mirrors, two beams of light recombine with each other at the beamsplitter.

If the distances travelled by two beams are the same which means the distances between two mirrors and the beamsplitter are the same, the situation is defined as zero path difference (ZPD). But imagine if the movable mirror moves away from the beamsplitter, the light beam which strikes the movable mirror will travel a longer distance than the light beam which strikes the stationary mirror. The distance which the movable mirror is away from the ZPD is defined as the mirror displacement and is represented by ∆. It is obvious that the extra distance travelled by the light which strikes the movable mirror is 2∆. The extra distance is defined as the optical path difference (OPD) and is represented by delta. Therefore, δ=2Δ

It is well established that when OPD is the multiples of the wavelength, constructive interference occurs because crests overlap with crests, troughs with troughs. As a result, a maximum intensity signal is observed by the detector. This situation can be described by the following equation: δ=nλ with n = 0,1,2,3...

In contrast, when OPD is the half wavelength or half wavelength add multiples of wavelength, destructive interference occurs because crests overlap with troughs. Consequently, a minimum intensity signal is observed by the detector. This situation can be described by the following equation: δ=(n+1/2)λ with n = 0,1,2,3...

These two situations are two extreme situations. If the OPD is neither n- fold wavelengths nor (n+1/2)-fold wavelengths, the interference should be between constructive and destructive. So the intensity of the signal should be between maximum and minimum. Since the mirror moves back and forth, the intensity of the signal increases and decreases which gives rise to a cosine wave. The plot is defined as an interferogram. When detecting the radiation of a broad band source rather than a single-wavelength source, a peak at ZPD is found in the interferogram. At the other distance scanned, the signal decays quickly since the mirror moves back and forth. Figure 4.14 shows an interferogram of a broad band source.

Figure 4.14 : Process in a Michelson Interferometer

Source:-https://slideplayer.com/slide/5905004/

• If M1 = M2, there is no optical path difference (OPD), and hence radiation is recombined at the beam splitter in phase (constructive interference). • If M1 ≠ M2, radiation recombining at the beam splitter is phase shifted and may undergo destructive interference (depending on wavelength)

Figure 4.15 : Explanation of the Michelson’s Interferometer - Source:https://slideplayer.com/slide/5905004/

4.6) Fourier Transform of Interferogram to Spectrum

The Michelson Interferometer uses interference to generate a spectrum (variation of intensity with wavelength for a source of a EM radiation) – used widely in remote sensing , particularly in Infrared.

Since spectrometers are equipped with polychromatic light source ( many wavelengths) the interference already mentioned occurs at each wavelength as shown in the lower figure. The interference pattern

74 produced by each wavelength are summed to get the resulting interferogram.

The interferogram is a function of time and the values outputted by this function of time are said to make up the time domain. The time domain is Fourier transformed to get a frequency domain, which is deconvolved to product a spectrum. Figure 4.20 shows the Fast Fourier transform from an interferogram to its spectrum.

The end result in an Interferogram. An Interferogram has every infrared frequency encoded in it. Characterized by a center burst at zero optical path difference and a complex pattern of waves around it. The interferogram is not recognized as a spectrum , because it is a function of time , whereas a spectrum is a function of frequence (or wavelength).

To produce a spectrum we compute the cosine Fourier Transform of the interferogram essentially a frequence analyzer extracting the component frequences of the interferogram.

Figure 4.16: Intensity-frequencies www.brukeroptics.com

Figure 4.17: Intensity – optical retardation ( www.brukeroptics.com )

Figure 4.18 : Resulting detector signal ( www.brukeroptics.com )

Figure 4.19 : Resulting interferogram  detector signal after modulation by a Michelson interferogram(source brukeroptics.com)

4.7) Extracting the spectrum from raw data It is predictable that the raw data collected on a Fourier transform spectrometer will be quite difficult to read. A Fourier transform needs to be performed to decode interferogram and extract actual spectrum I(v)

76 from it. The following shows how to conduct a Fourier transform to decode:

The intensity collected by the detector is a function of the path length differences in the interferometer p and wavenumber :

Thus, the total intensity measured at a certain optical path length difference (for each data point at a certain optical path-length difference p) is:

∫

It shows that they have a cosine Fourier transform relationship. So by computing an inverse Fourier transform, we can resolve the desired spectrum in terms of the measured raw data I(p) :

∫ ( )

Figure 4.20 : (a) Interferogram of a polychromatic light ; (b) its spectrum - (Source : BrukerOptics.com)

4.8) The Fast Fourier Transform (FFT) Fast Fourier Transform (FFT) is a very efficient algorithm to compute Fourier transform. It applies to Discrete Fourier Transform (DFT) and its inverse transform. DFT is a method that decomposes a sequence of signals into a series of components with different frequency or time intervals. This operation is useful in many fields, but in most cases computing it directly from definition is too slow to be practical. Fast Fourier Transform algorithm can help to reduce DFT computation time by several orders of magnitude without losing the accuracy of the result. This benefit becomes more significant when the number of the components is very large. FFT is considered a huge improvement to make many DFT-based algorithms practical. In Fourier transform spectrometer, signals are often collected by a series of optical or digital channels at the detector. Then FFT is of great importance to quickly achieve the following signal processing and data extraction based on DFT method.

Fourier transform spectroscopy can be applied to a variety of regions of spectroscopy and it continues to grow in application and utilization including optical spectroscopy, infrared spectroscopy (IR), nuclear magnetic resonance, electron paramagnetic resonance spectroscopy, mass spectrometry, and magnetic resonance spectroscopic imaging (MRSI). Among them, Fourier Transform Infrared Spectroscopy (FTIR) has been most intensively developed, which uses scanning Fourier transform to measure the mid-IR absorption spectra. .

Figure 4.21 : How FT transform the signal from the time domain to its representation in the frequency domain ( source - https://www.allaboutcircuits.com/technical-articles/an-introduction-to-the-discrete- fourier-transform/)

4.9) Background Spectrum If the interferometer chamber is not evacuated or purged with dry gas some absorption from the atmospheric Co2 and H20 is observed. This is called the background spectrum.

When the sample is introduced the spectrum now is a superposition of the absorption bands of the sample on an uneven background. To obtain %T with wavenumber we ratio the single beam sample spectrum with the background spectrum. Instead of analysing each spectral component sequentially, as in a dispersive IR spectrometer, all frequencies are examined simultaneously in FTIR spectroscopy. By interpreting the infrared absorption spectrum, information about the structure and the nature of the chemical bonds in a molecule can be determined.

The IR spectrum is usually displayed as % transmittance, that is how much of the original IR intensity is left after passing through the sample. In order to calculate this, we first need to know the IR intensity with no sample, so we run a background.

The first thing you notice about the background spectrum is there are strong IR peaks near 3800, 2400 , 1600 cm-1. These peaks are due to the

O-H stretch of H2O, the asymmetric stretch of CO2 and the H-O-H bending of H2O, respectively. The H2O and CO2 are present in the air, which fills the spectrometer. The H2O bands consist of many sharp peaks. These peaks are the vibrational transitions between different rotational states of H2O. With a higher resolution instrument, the CO2 band shows similar rotational splitting, but the lines are closer together.

Even without the CO2 and H2O bands, the background spectrum is much more intense in the middle than at both ends reflecting the output spectrum of the source: strong in the middle, but falling off at the ends.

Figure 4.22 : A reference single channel ( Thessaloniki – 10-07-2019 )

4.10) Absorption spectroscopy

Absorption spectroscopy refers to spectroscopic techniques that measure the absorption of radiation, as a function of frequency or wavelength, due to its interaction with a sample. The sample absorbs energy, i.e., photons, from the radiating field. The intensity of the absorption varies as a function of frequency, and this variation is the absorption spectrum. Absorption spectroscopy is performed across the electromagnetic spectrum.

Absorption and transmission spectra represent equivalent information and one can be calculated from the other through a mathematical transformation. A transmission spectrum will have its maximum intensities at wavelengths where the absorption is weakest because more light is transmitted through the sample. An absorption spectrum will have its maximum intensities at wavelengths where the absorption is strongest.

4.11) Transmission spectrum

The single channel sample spectrum is different from the background one. It shows less intensity at those wavenumbers where the sample absorbs radiation.

Here we appose a theoretical example of a spectrum from ( Brukeroptics.com ) and we compare it with our experimental results for the region of Thessaloniki.

Figure 4.23 : Single channel intensity for sample and reference spectrum to wavenumber

Pictures are taken by Brukeroptics.com

T(v) = S(v) / R(v)

Where S(v) is the sample spectrum and R(v) is the reference one.

Figure 4.24: Spectral region ( 4000 – 15797 cm )  one measurement (10 scans , 58 seconds)

Furthermore, we can check our spectra to realize if a specific day was cloudy. If the intensity of the single channel is reduced dramatically, then it means that the absorbance of radiation was very high and subsequently it shows less intensity in all wavenumbers.

To understand this ,we appose here two different cases where in the first one we have also absorption from clouds, while in the second one we observe an absolutely clear sky where the peaks represent the spectral regions that every x-gas absorb the incoming radiation.

Figure 4.25 : The reduction in radiation which is being received can be clearly observed (taken from our measurements in Thessaloniki (2019/07/09)

In this part I would like to point out that the x axis represents the wavenumbers and the y axis is arbitrary units which represents a relative unit of measurement to show the ratio of amount of substance, intensity or other quantities to a predetermined reference measurement. The reference measurement is typically defined by the local laboratories or dependent on individual measurement apparatus.

Figure 4.26 : Experimental FTIR measurements in Thessaloniki (19/06/14) in a –clear sky- day

Gas Spectral bound ( wavenumbers ) Spectral First Second windows XCO2 6173.00-6390.00 XCH4 5897.00-6145.00 4208.7-4318.8 XCO 4208.7-4318.8 (CO) XH2O 8353.4-8463.1 XO2 7765.0-8005.0 XAIR 7765.0-8005.0

Table 4.1.2 : Spectral micro-windows for X- gases and X-air

In the next few graphs we present the absorption lines of these x – gases in function of the wavenumbers which every gas shows less intensity in arbritary units , which means that molecules excites in higher energy states. Absorption peaks in an infrared absorption spectrum arise from molecular vibrations. Absorbed energy causes molecular motions which create a net change in the dipole moment.

Figure 4.27 : XCO lines – wavenumber

Figure 4.28 : XCH4 lines – wavenumber

Figure 4.29 : XCO2 lines – wavenumber

Figure 4.30 : XO2 lines – wavenumber

Figure 4.31 : XH2O lines – wavenumber

5) Analysis of FTIR ground-based measurements

5.1) CamTracker program

The Camtracker consists of a homemade altazimuthal solar tracker, a digital camera and a homemade program to process the camera data and to control the motion of the tracker. The key idea is to evaluate the image of the radiation source on the entrance field stop of the spectrometer.

Figure 5.1 : CamTracker (EM27/SUN operation manual – Bruker.com)

The tracking mechanism is based on astronomical calculations for a coarse tracking and an optical feedback provided by a camera.

It is based on the publication: M. Gisi, F. Hase, S. Dohe and T. Blumenstock: Camtracker: a new camera controlled high precision solar tracker system for FTIR-spectrometers, Atmos. Meas. Tech., 4, 47-54, doi:10.5194/amt-4-47-2011,2011 , http://www.atmos-meas tech.net/4/47/2011/amt-4-47-2011.pdf Bruker Optics also applied a patent for this principle.

In the ideal case, the solar image is centered on the field stop opening of the spectrometer. For this reason, a camera is built into the spectrometer, recording the field stop.The CamTracker program detects the rim of the solar image as well as the field stop opening at the same time, which, by principle, minimizes misalignments significantly. (Hase et al. 2011).

The rim of the sun is detected by setting a brightness threshold to the image using the brightness histogram to separate the illuminated and non-illuminated areas.

Figure 5.2 : Brightness histogram of the camera image. The huge number on low illuminated pixels on the left side corresponds to the non-illuminated field stop regions (field stop opening and area around the solar image). The horizontal line marks a selected threshold., http://www.atmos- meas-tech.net/4/47/2011/amt-4-47-2011.pdf , Bruker Optics

The operation principle of our tracker is based on a combination of astronomical algorithms to provide the coarse mirror angles with a superposition of small corrections to these angles derived from the optical feedback providedby the camera. The recorded pictures are evaluated byour software “CamTrack” in real-time, in order to determine both the central position of the field stop opening and the solar disk. Then the necessary mirror angle corrections to“move” the solar disk on the field stop wheel to the desired position relative to the input field stop opening are calculatedand sent to the tracking system. These steps are continuously performed about three times per second. If no usable positions for the solar disk and the field stop opening can be

90 determined, for example due to clouds, the system continues tracking on the basis of the astronomical coordinates together with previously saved offset values for similar tracking angles. (Hase et al. 2011).

The main steps in determining the central positions from the camera pictures are the following:

 Finding an appropriate threshold value to separate the bright area illuminated by the sun from the dark, nonilluminated rest of the field stop wheel and the dark opening of the field stop.  2. Creating a binary (black/white) picture by applying the threshold, and finding the contours along the obtained areas (solar disk and field stop).  3. Fitting ellipses along the contours (in a least squares sense).  4. Performing consistency checks, if the obtained ellipses can be the contours of the solar disk and the field stop opening in terms of criteria like radius, position and eccentricity.  5. If the previous step was successful, taking the centers of the ellipses as centers of the solar disk and the round field stop opening.

Where the field stop opening is not inside the solar disk, it is not visible. Then the solar disk is moved along a search pattern over the field stop wheel , until the opening is found.(Hase et al.)

As we can observe in the following pictures when clouds are across the sun and as a subsequence we have no incident radiation from the sun , the image is not appropriate for our measurements :

Figure 5.3 : Here we observe the deviation of the radiation of the sun while the tracker is trying to achieve the best fit in the field of view of the spectrometer

Figure 5.4 : Two pictures of the camera. The downside picture shows a correct positioning of the solar disk relative to the field stop opening despite strong intensity variations resulting from clouds.

Pictures are taken from FTIR measurements in Thessaloniki

5.2) Recording Spectra ( OPUS_7.2.139.1294 )

The spectra can be recorded using software called OPUS. We can specify many things inside this clever software such as the number of scans during a measurement , FFT , apodization function , phase error and a lot of other options.

First of all in the basic TAB we insert the path in which our measurements would be extracted.

In the advanced mode we specify resolution ( 0.5 cm^(-1) ) , number of the sample scans (10) , also for the background spectrum , our spectral coverage , result spectrum (TRANSMITTANCE) and data blocks to be saved.

In the next step (Optic) we set the scanner velocity at 10 KHz and in Acquisition laser wavelength ( 15798.10 ) , Interferogram size points , a high pass filter , a low pass (10KHz) and double sided Forward and Backward interferograms.

Last but not least ,we perform Phase Correction mode selecting Mertz option , which is intended to give phase error during the processing in Michelson’s Interferometer. After that , we have the opportunity to perform an Apodization Function ( Norton-Beer , Medium ) in order to eliminate side lobes right and left of the interferogram. Finally ,a zero- filling factor eighth grade is being used.

The most important thing in this mode is to save our peak position ( Check Signal ) before we start our measurements (RunMacro) .

5.3) TCCON

The Total Carbon Column Observing Network (TCCON) is a network of ground-based Fourier Transform Spectrometers that record spectra of the sun in the near-infrared. From these spectra, accurate and precise column-averaged abundances of atmospheric constituents including

CO2, CH4, N2O, HF, CO, H2O, and HDO, are retrieved.

The TCCON was established in 2004 to retrieve precise and accurate total column amounts of atmospheric trace gases like CO2, CH4, N2O and CO from ground-based solar absorption spectra in the NIR region [Toon et al., 2009] [Wunch et al., 2010] [Wunch et al., 2011]. These observations rely on spectroscopic parameters that are not known with sufficient accuracy to compute total columns that can be used in combination with in situ measurements. The TCCON must therefore be calibrated to World Meteorological Organization (WMO) in situ trace gas measurement scales. To derive a total column measurement of the gases from these spectra, external information about the atmosphere (e.g. temperature, pressure, a priori mixing ratio) and NIR spectroscopy is required. Meanwhile there are 19 TCCON stations worldwide, with gaps in South America, Africa and Asia. Some of the TCCON sites have in situ measurements nearby, which can be used for comparison and calibration.

Source : https://en.wikipedia.org/wiki/Total_Carbon_Column_Observing_Network

Figure 5.5: TCCON sites around the globe, as of March 2017

Source: https://en.wikipedia.org/wiki/Total_Carbon_Column_Observing_Network

5.4) The HITRAN Database

HITRAN is an acronym for high-resolution transmission molecular absorption database. HITRAN is a compilation of spectroscopic parameters that a variety of computer codes use to predict and simulate the transmission and emission of light in the atmosphere.

HITRAN is maintained and developed at the Harvard-Smithsonian Center for Astrophysics, Cambridge MA, USA.

HITRAN is the worldwide standard for calculating or simulating atmospheric molecular transmission and radiance from the microwave through ultraviolet region of the spectrum. The current version contains 49 molecular species along with their most significant isotopologues. These data are archived as a multitude of high-resolution line transitions, each containing many spectral parameters required for high- resolution simulations. In addition there are 320 molecular species collected as cross-section data. These latter include anthropogenic

95 constituents in the atmosphere such as the chlorofluorocarbons. (https://www.cfa.harvard.edu/hitran/)

The simultaneous developments of high-resolution laboratory instrumentation (such as the Fourier transform spectrometer-FTIR), the digital computer and storage, and sensitive detectors and the means to carry them on board high-altitude balloons and space craft provided the stimulus to create a machine-readable archive of the fundamental properties of molecular transitions. It was then possible to simulate transmission and radiance in the terrestrial atmosphere by applying known radiative-transfer equations. Thus was born the original HITRAN molecular absorption line parameters database. (https://www.cfa.harvard.edu/hitran/)

The initial HITRAN was limited to the seven main telluric atmospheric absorbers in the infrared: H2O, CO2, O3, N2O, CO, CH4, and O2. The most significant of the isotopologues of these molecular species was also included. The initial HITRAN database included only the basic parameters necessary to solve the Lambert-Beers law of transmission, namely the line center of a transition, the intensity of the transition, and the lower-state energy. In addition, the air-broadened Lorentz width was included as well as the unique quantum identifications of the upper and lower states of each transition.

The experimental data that enter HITRAN often come from the results of analysis of Fourier transform spectrometer laboratory experiments. Many other experimental data also are inputted, including lab results from tunable-diode lasers, cavity-ring down spectroscopy, heterodyne lasers, etc. The results usually go through elaborate fitting procedures. The theoretical inputs include standard solutions of Hamiltonians, calculations, and semi-empirical fits.

Another aspect of HITRAN is that when line-by-line transition parameters are either not available, or not practical, experimentally-

96 determined cross-sections at different temperatures and pressures are provided. This part of HITRAN can be used for quasi-physical radiative- transfer solutions. Examples of these molecules are anthropogenic gases including the chlorofluorocarbons, large molecules, etc. (https://www.cfa.harvard.edu/hitran/)

Figure 5.6 : Hitran.org (https://www.cfa.harvard.edu/hitran/)

5.5) COCCON software description, version: 180806

First of all, we want to thank KIT for providing us with this useful code for the retrieval of x-gases!

This documentation explains in short the usage of software tools provided here for the COCCON (Collaborative Carbon Column Observing Network) data analysis. The development of the preprocessing tool has been supported by ESA. All three codes are written in Fortran and are largely portable. The preprocessing tool is fully compatible with Fortran 2003 standard and therefore is fully portable to other platforms. The “pcxs” and “invers” codes follow Fortran 95 standard using a few Lahey compiler specific non-standard commands, so minor adjustments might be required when using other compilers or operating systems.

The codes have been tested on a large dataset of raw measurements and for different EM27/SUN spectrometers.

The data processing has two steps, reflected in 1 + 2 independent pieces of code which are:

 The preprocessing: generates spectra in a binary format which are then used for the subsequent quantitative trace gas analysis. This code also checks the quality of the raw data, it generates a *.bin output spectrum only if no problems were detected during the processing chain. It performs a DC correction, FFT, phase correction, and a resampling of the spectrum to a minimally sampled grid. This code has been created by KIT in the framework of ESA’s COCCON-PROCEEDS project.

 The quantitative trace gas analysis. This is a 2-step-sequence: (1) The “pcxs” program (pcxs: precalculate x-sections) is used for tabulating the cross-sections of all relevant gases in the ~4000- 8600 cm-1 spectral region. Here we have to infer that the pressure values taken by the observer in the observer altitude (intraday file) , are being taken every 10 or 15 minutes. (2) afterwards, the

“invers” program (performs the inversion) is used for performing the trace gas retrievals for all species on all available calibrated spectra. The program fits the column amount of each target gas (and further interfering species and auxiliary quantities) and finally performs the required post-processing steps for creating the final

estimates of column-averaged abundances (called XCO2, XCH4, etc) by applying airmass-independent and airmass-dependent corrections.

Here I appose the methodology that we applied , which has been described above.

5.5.1) Process In the pre-processing step there is a processing loop for all our spectra during the measurement day. Our smart code is reading the file from the disc , performs a DC correction and then preparing the Fast Fourier Transform (FFT as we mentioned before in our thesis) .

We insert our latitude , longitude as well as the Instrumental Line Shape (ILS) – we will explain later in our thesis what is ILS . The spectra are named HHMMSSSN.bin (main channel) or HHMMSSSM.bin (CO channel). The time used for the naming is the start time of the measurement as given in the OPUS file, so we adopt our choice of time zone when the OPUS file is generated.

To continue, we are obliged to do is to ‘run’ preprocess4.exe to get our spectra in order to use them to retrieve the total columns of our gases in the next steps of our code.

The next step ( pre-calculation of x-sections ) requires location and date- specific meteorological information: the atmospheric pressure and temperature profiles and the a-priori mixing ratio of each trace gas (H2O, HDO, CO2, N2O, CO, CH4, O2, and HF). In order to be compatible

99 with the TCCON analysis, we directly process *.map files which we receive from KIT University.

The observer’s altitude,longitude,latitude and atmospheric pressure is provided and given as well as the solar spectrum file from map file.

After these actions line datas are being reading from HITRAN for each of these gases. 49 levels of altitude are given concluded the one from the observer and solar spectrum is reading , interpolating and is being written to file. So we have the profiles after reading the pT file and VMR files so that the total columns to be retrieved.

Furthermore, as long as reading line data of HITRAN is completed the tau_levels for species are given by calculating total optical thickness , air mass dependence factor (AMDF).

For each gas a polynomial approach is proceeded to extract the levels in function of temperature and pressure in order to calculate the total columns by interpolating the map files.

100

The Intraday txt file has the values of pressure taken by the observer using the meteorological station in AUTH and we would like to thank the meteorological department (https://meteo.geo.auth.gr/) for their support and help by allowing us to have access to these valuable data.

101

Above pictures represent the pre-process step and the pre-calculation of x sections.

5.5.2) Starting the retrieval After the x-section table has been generated, the inversion of all spectra recorded during a single measurement day can be performed. While the

102 pre-calculation of x-sections needs to assume a single set of daily meteorological parameters, it also provides derivatives of x-sections with respect to ground pressure and boundary layer temperature. Therefore, intraday ground pressure changes and intraday changes of PBL temperature due to radiative heating of the ground can be taken into account. The assumed PBL thickness is already set in “pcxs.inp”: the n_Tdisturb variable sets how many levels are affected by a change of temperature. Typically, the PBL will heat up and thicken during the day, the latter effect is neglected, but the temperature change could be scaled to approximate this effect. Note that the relevant temperature change strongly differs the temperature change measured at the observer altitude, the amplitude will be much lower than the temperature change at ground. After OPDmax, apodization function ,phase error is given , the calculation for each gas in specified spectral bounds is getting started. A polynomial fourth grated, the observer altitude and pressure as well as a CO spectra is also taken into account. For each spectrum an intraday pressure and temperature is calculated for every available level of height.

The post processing includes airmass independent and airmass dependent corrections. These corrections are set according to the values provided in the invariable part of the “invers.inp” input file. These values have been adjusted to bring the COCCON results in close agreement with TCCON. Exceptions are XCO, where we assume that the TCCON calibration has a low bias of about 5%, and the deviating definition of XAIR: here we use the ratio of the spectroscopically derived ground pressure over the measured ground pressure, and we apply a normalization factor to ensure the value is close to unity for an ideal measurement. XAIR is a valuable indicator for instrumental problems, time offsets, incorrect observer coordinates, and incorrect values for pressure at observer altitude.

103

Inversion of all spectra recorded during a single measurement day

104

6) Misalignments

 X-air - introduction

Xair compares the measured oxygen column (VCO2 ) with surface pressure measurements (PS)

Here μ and μH2O denote the molecular masses of dry air and water vapour, respectively, g is the column averaged gravitational acceleration and VCH2O is the total column of water vapour. The correction with VCH2O is necessary as the surface pressure instruments measure the pressure of the total air column, including water vapour. For an ideal measurement and retrieval with accurate O2 and H2O spectroscopy, as well as accurate surface pressure, Xair would be 1. However, due to insufficiencies in the oxygen spectroscopy, this value is not obtained . For the EM27/SUN prior studies showed a factor of 0.97 (Frey et al., 2015; Hase et al., 2015; Klappenbach et al., 2015). Large deviations ( 2 %) from these values indicate severe problems, e.g. errors with the surface pressure, pointing errors, timing errors or changes in the optical alignment of the instrument(Frey et al., 2015; Hase et al., 2015; Klappenbach et al., 2015).

In this work the X-air shows a deviation of 1.9-2.6 % due to uncertainties in the ground pressure.

Measurements of total column abundances of trace gases in the atmosphere are subject to a variety of restrictions of the remote-sensing technique. Several sources of uncertainty have to be taken into account

105 which impact the accuracy of retrieved abundances from solar absorption spectra. Some errors are considered as systematic (e.g. spectroscopic parameters, detector nonlinearity, instrumental changes over time like ILS drifts) while others are of statistical behaviour (e.g. spectral noise). Errors due to solar spectral photon noise are in general small for instruments with a direct solar viewing geometry. Moreover, most of the spectral windows considered by the NDACC and TCCON retrieval strategies contain strong spectral absorption lines. Spectral noise has a liberal effect on retrieved column abundances since it causes no significant bias. Assuming a retrieval strategy which performs a scaling retrieval, the spectral noise error is estimated via:

√

Here, 푆 represents the line intensity of the observed spectral absorption line, 휎 the standard deviation of the spectral noise and 𝑁 the number of independent spectral grid points within the half width of the spectral line (Hase, 2000). The error which affects the retrieved total column is of the order of 0.03 % for CO2 (Wunch et al., 2011; Dohe, 2013). Moreover, the signal-to-noise ratio (SNR) for a single spectrum (75 sec acquisition) recorded at 45 cm OPDmax is approximately 750 : 1 in the region of 5000 cm−1. https://chem.libretexts.org/Courses

6.1) Ghost to parent ratio

Ghosts are artificial spectral features linked to the aliasing of true spectral lines that arise in FTS spectra.

The InGaAs detectors are optically sensitive at wavenumbers greater than half the HeNe metrology laser frequency (7899 cm^-1). The ghost to parent ratio is 1.73*10^-4 at a 10 kHz acquisition rate without the interpolated sampling activated. The sampling of the interferogram has to be performed at every zero crossing of the laser signal (HeNe laser, wavelength 633 nm). If the signal

106 is not taken at exactly zero intensity, systematic sampling errors are introduced leading to artifacts in the measured spectrum (so-called sampling ghosts). https://chem.libretexts.org/Courses

The Brownian motion of the molecules in the atmosphere has two main consequences: pressure broadening and Doppler broadening.

6.2) Pressure broadening

Pressure broadening occurs due to collisions of molecules, causing a shortened lifetime of excited states and broadened line shapes. The higher the number density of molecules, the larger is the probability of a collision. This means that pressure broadening has a larger effect on lower atmospheric levels than on higher ones. The resulting lines have, just like the natural broadened ones, a Lorentzian shape :

√

Platt and Stutz (2008)

6.3) Doppler broadening

Doppler broadening occurs due to the Doppler effect caused by the thermic motion of molecules. If such a molecule moves with velocity in the observer's direction and emits radiation with the angular frequency ωο, the radiation reaching the observer is then shifted to ω:

107

The Doppler broadening is dependent on temperature as well as the molecular mass and is proportional to the wavenumber.

For a typical atmospheric scene both pressure and Doppler broadening contribute to the shape of a spectral line. The Voigt profile is the convolution of the Gaussian and Lorentzian profile: Platt and Stutz (2008)

( ) ∫ ( ) ( )

Figure 6.1 : Sketch of the different types of line profiles: Lorentz shape (collision broadening), Doppler (Gaussian) shape (thermal motion of molecules), and Voigt shape resulting from simultaneous Doppler and collision broadening. Platt and Stutz (2008).

6.4) Air – mass dependency

Another important error source is the solar zenith angle (SZA), which has to be well known to determine the air mass between the light source (Sun) and the spectrometer.

108

Considering a tracking range , the air mass changes by up to 9.6% per degree SZA change.

The actual observed air mass differs from the air mass assumed in the analysis. The error resulting from a line-of-sight (LOS) error depends strongly on the zenith angle of the sun.

SZA angle dependency :

6.5) ILS

Precise knowledge of a spectrometer’s instrumental line shape (ILS) is of utmost importance to gain correct information from measurements because using wrong ILS values leads to systematic errors in the gas retrieval. The ILS can be divided into two parts. One part describes the modulation loss through inherent self-apodization of the spectrometer which is present also in an ideal instrument. This contribution can easily be calculated utilizing the OPD and FOV of the spectrometer. The other component of the ILS results from misalignments and optical aberrations of the spectrometer and can be characterized by a modulation efficiency amplitude and a phase error, both functions of the OPD (Hase et al., 1999). These parameters have to be deduced from lab measurements.

Due to the limited path of the mirror, the interferogram is truncated at OPD max, leading to spectral artifacts which have to be eliminated. The interferogram's truncation can be described mathematically as the multiplication of an infinite interferogram Ioo (x) with a boxcar function B (x):

109

The measured spectrum I (~) is the convolution of the Fourier transforms of Ioo (x) and B (x), where the Fourier transform C (v) of B (x) is the following sinc - function:

The width of the sinc-function denes the spectral resolution, where the full width at half maximum (FWHM) of C (v) is :

The sinc-function is also called the instrumental line shape (ILS), being a measure of how the spectrometer deforms a sharp spectral line. Due to misalignments, the real ILS might be asymmetric or might have a smaller amplitude than the ideal one, which has to be considered within the retrieval. We obtain the final spectrum used in the subsequent analysis by taking the averaged interferogram, performing a DC-correction and a Fourier transformation. In order to predict the correct width of the observed H2O lines, thereby correctly retrieving the ILS width, the distance between instrument, measured from the first tracking mirror, and lamp needs to be measured as well as air temperature and pressure at the time of measurement. (M. Frey et al. 2015)

In a few words …

ME and PE both describe the interferogram and vary with OPD.

 PE is the angle between the real and imaginary parts of the FT of the ILS.

110

 ME is a measure of the normalized observed interferogram signal compared with that of a nominal instrument with an ideal value of 1 • At maximum OPD (MOPD), an ME< 1 causes a broadening of the measured spectral lines, while an ME> 1 at MOPD causes a narrowing.

The ILS can be calculated by analyzing absorption lines measured through a low-pressure gas cell, and varies with OPD. https://chem.libretexts.org/Courses

6.6) Zero filling

Zero filling is a data processing technique where zero points are added in the end of the Interferogram before the digital Fourier Transformation. It is the process of interpolating extra data points into a spectrum so that the spectral lines have a smoother shape with better digital resolution , using the same Fourier coefficients. FTIR software automatically provides zero-filling by extending the length of an interferogram with a zero straight line.

Zero filling is an interpolation that does not affect the instrument line shape , and in most cases , is therefore superior to polynomial or spline interpolation methods that are applied in the spectral domain. (source : brukeroptics manual )

111

Figure 6.2 : In our measurements in Thessaloniki we apply a zero – filling factor eighth grade ( picture taken by Bruker-Optics )

6.7) Apodization Function

A transmittance spectrum (or a spectrum converted to an absorbance spectrum) is obtained when Fourier transform is applied to the measured interferogram. In particular, the integration range is from 0 to infinity. This supports an infinite range of movement of the moving mirror. However, such a movement is impossible. The moving mirror reciprocates through a finite distance, such that in practice this integration has to be cut off in a finite range. ( source : brukeroptics manual )

In a real measurement, the interferogram can only be measured for a finite distance of mirror travel. The resulting interferogram can be thought of as an infinite length interferogram multiplied by a boxcar

112 function that is equal to 1 in the range of measurement and 0 elsewhere. This sudden truncation of the interferogram leads to a sinc(v) instrumental line shape as we mentioned above, the meaning of ILS. For an infinitely narrow spectral line the peak shape is shown at the top of the figure. The oscillations around the base of the peak are referred to as ‘ringing’ or ‘leakage’. The solution to leakage problem is to truncate the interferogram less abruptly. This can be achieved by multiplying the interferogram by a function that is equal to 1 at the center-burst and close to 0 at the end of the interferogram. This is called APODIZATION and the simplest one is a rump or triangular apodization factor. There are many options to select the appropriate apodization function in our interferogram for the purpose that we need this formation. (source : brukeroptics manual )

In our work , we selected the Norton-Beer Medium apodization function !!

Figure 6.3 : Multiple choices of selecting the appropriate apodization function for our work (source brukeroptics.com)

113

Figure 6.4 : Some common apodization functions and the resulting line shapes (brukerOptics manual)

6.8) Phase Error

The signal processing that converts the interferogram into a spectrum usually involves phase correction, zero-filling and apodization as well as the Fourier transform. An ideal interferogram would be perfectly symmetrical about zero path difference. However, in practice there is some asymmetry because of phase differences between the signals from different wavenumbers. A phase correction routine is applied before the Fourier transform. Phase correction can be avoided by scanning the full length of the interferogram on both sides of zero path difference. This allows a magnitude spectrum to be calculated without phase correction. (source : Brukeroptics manual )

In Fourier transform spectroscopy interferograms are apodized prior to the transformation and the calculated spectrum is phase corrected. If the zero point for the Fourier transformation is displaced, a phase error occurs, because the apodization remains fixed while the interferogram is shifted by the phase correction. To avoid distortions of the spectrum resulting from phase errors, double-sided interferograms are needed. In practice, single-sided interferograms are recorded double-sided on a small range of optical path difference, allowing determination of the phase error.

114

In such a case, the phase error is determined at low resolution, typically about 1 cm^(-1) or less. This is not a problem because it varies only slowly with wavenumber. These phase shifts lead to an asymmetry in the interferogram. This asymmetry is corrected during the DFT using one of a number of phase corrections algorithms. (source : brukeroptics manual )

The algorithm that we use in our study was developed by Larry Mertz and is therefore called “ Mertz Phase Correction “ .

Figure 6.5 : Michelson’s Interferometer-(Source : Brukeroptics.com)

115

7) Ground-based FTIR measurements in Thessaloniki 2019

7.1) Set up

The EM27/SUN was place on the roof of the Laboratory of Atmospheric Physics (AUTH) and our measurements were started from 15th of January in the morning , in a very windy day but also sunny. This setup achieves high stability against thermal influences and vibrations. Gimbal-mounted retro-reflectors move a geometrical distance of 0.45 cm leading to an optical path difference (OPD) of 1.8 cm which corresponds to a spectral resolution of 0.5 cm^(-1).

The modulated signal is detected by an InGaAs detector covering the spectral domain from 5000 to 11000 cm^(-1) and is called an interferogram. As the EM27/SUN analyses solar radiation, it can only operate in sunny daylight conditions. A Fourier transform of the interferogram generates the spectrum and a DC correction is applied to remove the background signal and only keep the AC signal (Keppel-Aleks et al., 2007). The detector is a photodiode with a size of 1mm_1mm and spectral coverage from 4500 to 12000 cm^(-1). In contrast to the detector used in the prototype that operated in the spectral region between 6000 and 9000 cm^(-1), the wider spectral coverage allows the observation of CH4.

We orientated the instrument carefully in order to avoid unwanted distortions and vibrations towards to the south so that the tracker could be able to follow the movement of the sun during the day. Also , we set our latitude , longitude and altitude in order to get the right coordinates. In a first step we started the initialization of the tracker. Then we checked the sun in the processing mode and the tracker began to target the sun. Because of not so great orientation was achieved we manually

116 rotate the spectrometer in order to be satisfied with the entrance of the laser beam into the field of view. Then, we activated the camera to detect the actual line of sight of the spectrometer and to correct the solar tracker angles. So we were ready to start our measurements in OPUS (as we have mentioned above in this work in details), while the correct peak position in the interferometer was saved successfully.

For the analysis of “our” gases we don’t use the total column measurements , but rather we use X-gases as misalignments such as ILS errors or tracking errors ( as we mentioned above in the previous section ) mostly cancel out.

But lets see how do we measure X – gases :

 O2 as an internal standard: calculation of dry-air mole fractions

The column abundances of gas G are converted to column-averaged DMFs by dividing them by the total column of dry air:

This division removes variations due to surface pressure changes, making results from different days or sites more directly comparable. We could compute the column of dry air using a surface pressure measurement and a water vapour measurement:

{ } where mH2O and mdry air are the mean molecular masses of water and dry air, respectively, Ps is the surface pressure and {g}air is the column- averaged gravitational acceleration. An alternative method of measuring the column of dry air is to measure

117 the column of O2 from the FTS spectrum, and divide by the known DMF of O2 (0.2095):

This approach has several advantages. Errors that are common to the target gas (e.g. CO2) and O2 bands (such as miss-pointing or zero-level offsets) will generally cancel in the column ratio. Another advantage is that the water column is not required to correct the surface pressure. The O2 column can also be used as a network-wide internal standard, in lieu of standard gas mixtures that are used for the global in situ network, since, to the degree required here, the mixing ratio of O2 is constant and well known. Hence, for TCCON, the column of dry air is retrieved from the O2 column, and DMFs are computed using :

Source : Hase et al. 2012

7.2 ) Mean daily values and daily course of FTIR ground-based measurements in Thessaloniki

In Table 7.2 mean daily values ( in part per million {ppm} ) of each x-gas from January 15th until August 1st are shown , as well as the graphs for these gases where in y-axis we have the mean value and in x-axis date is being shown.

Furthermore , we present the daily course of some measurement days in Thessaloniki for every x-gas and we try to understand why peaks arise in some figures and which sources are responsible for.

118

MEAN VALUES OF X-GASES Dates xco2_M xco_ME xh2o_ME xo2_MEA xch4_MEAN (YYMM EAN AN AN N (ppm) DD) (ppm) (ppm) (ppm) (ppm)

190115 402.50083 0.08830 609.23858 482.24242 1.81867 190214 406.43901 0.10982 1197.40826 957.95808 1.82525 190218 404.25980 0.09153 1631.21653 1333.10528 1.81257 190219 405.04276 0.09651 1842.99675 1471.66732 1.81316 190220 405.54386 0.10086 1538.49685 1219.52065 1.81342 190221 405.34135 0.10810 1135.37034 910.46545 1.82008 190318 405.17553 0.09722 2534.10833 2011.25890 1.80394 190329 406.87984 0.11730 1408.44221 1096.73471 1.82404 190417 406.44483 0.10707 1442.40566 1135.17129 1.81049 190506 407.94931 0.10132 2376.95906 1931.39031 1.81241 190508 406.06684 0.10155 1775.95226 1402.23000 1.81463 190512 407.16904 0.10761 3674.78588 2930.01868 1.82087 190521 406.63451 0.09502 2509.64655 2008.81586 1.81103 190523 406.01001 0.10116 4181.13163 3374.84921 1.82338 190528 406.98491 0.09824 4086.99847 3262.10851 1.82462 190529 407.34505 0.09692 4176.06506 3343.88815 1.81674 190603 405.78218 0.08844 4060.04904 3270.05246 1.80528 190604 404.86452 0.09687 4570.46864 3684.30804 1.82200 190607 406.36532 0.10192 5087.97446 4115.79609 1.80875 190611 404.62987 0.10119 5774.96056 4688.17071 1.82337 190613 405.36214 0.09871 4929.07482 3967.84625 1.81278 190614 406.79547 0.09805 5431.41656 4422.40711 1.81600 190620 404.57486 0.10342 6608.39480 5327.51677 1.82340 190621 404.95578 0.09538 5462.22138 4438.42841 1.82203 190627 402.84084 0.09658 5168.10259 4142.84844 1.81850 190628 403.92939 0.09015 5177.23833 4220.21114 1.81943 190701 405.72058 0.09210 4615.33648 3641.05231 1.80732 190702 405.71370 0.08408 4174.74229 3298.77865 1.80138 190703 407.21484 0.08626 4155.31700 3294.27730 1.80597 190704 406.63218 0.08519 4974.55788 3936.93061 1.79801 190705 405.91420 0.09092 5395.24322 4307.85541 1.80605 190709 405.63351 0.09093 6471.57478 5227.91944 1.81103

119

190710 406.59902 0.08477 6280.15168 5103.78353 1.80152 190712 401.48753 0.08530 4175.94217 3310.93261 1.81163 190715 403.74950 0.08801 4581.14959 3730.59600 1.82386 190718 400.86075 0.08919 4330.94976 3474.98712 1.81976 190719 401.51264 0.09384 5436.37600 4298.66558 1.82463 190722 400.63469 0.10234 6148.83114 4910.74429 1.83600 190723 400.96500 0.09681 6518.38198 5258.73236 1.81488 190724 403.84424 0.08840 5707.36184 4627.45079 1.81504 190725 403.94007 0.09618 6025.99400 4731.07800 1.82953 190726 402.01381 0.09638 5353.81408 4214.39258 1.82562 190729 403.25271 0.08903 5937.25567 4801.31546 1.81794 190730 403.84760 0.09358 6750.88470 5467.99931 1.82245 190731 402.18703 0.09299 6117.58737 4945.75157 1.82695 190801 402.43070 0.09784 7054.04481 5732.39240 1.83418 Table 7.2 : Mean values of X-gases (ppm) from January 2019 until August 2019 for the region of Thessaloniki

In Figure 7.2.1 mean values of XCO2 are shown from 15 January until the end of August. There is a slightly increase in the winter and spring due to anthropogenic emissions such as electricity generation and transportation and fossil fuel use - of course because we are in the heart of the city center of Thessaloniki and also industrial processes. Electricity and heat generation is the economic sector that produces the largest amount of man-made carbon dioxide emissions. Although , a decrease can be seen as well in the summer due to photosynthesis (which is a process used by plants and other organisms to convert light energy into chemical energy that can later be released to fuel the organisms' activities. This chemical energy is stored in carbohydrate molecules, such as sugars, which are synthesized from carbon dioxide and water – hence the name photosynthesis, from the Greek φῶσ, phōs, "light", and ςφνκεςισ, synthesis, "putting together" ) and also for the reason that not so much electric power is consumed by the residents of the city. Deforestation has been responsible for the great majority of these emissions. Deforestation is the permanent removal of standing forests and is the most important type of land use change because its impact on greenhouse gas emissions. Trees act as a carbon sink. They remove carbon dioxide from the atmosphere via

120 photosynthesis. When forests are cleared to create farms or pastures, trees are cut down and either burnt or left to rot, which adds carbon dioxide to the atmosphere.

Another possible emission by sector is solid waste disposal.

A minimum value of 400.63469 ppmv can be observed ( Figure 7.2.1 ) in July and a maximum value of 407.94930 ppmv can be also observed in May , as a consequence of burning fossil fuels in this traffic area.

Figure 7.2.1 : XCO2 mean daily values in respect with dates - Thessaloniki

121

Figure 7.2.2 : XCO2 daily measurements in Thessaloniki ( FEBRUARY 2019 )

Figure 7.2.3 : XCO2 daily measurements in Thessaloniki ( JULY 2019 )

122

Figure 7.2.4 : XCO2 daily measurements in Thessaloniki ( AUGUST 2019 )

Figure 7.2.5 : Overall equation for the type of photosynthesis that occurs in plants (https://en.wikipedia.org/wiki/Photosynthesis)

123

Figure 7.2.6 : Composite image showing the global distribution of photosynthesis, including both oceanic phytoplankton and terrestrial vegetation. Dark red and blue-green indicate regions of high photosynthetic activity in the ocean and on land, respectively. (https://en.wikipedia.org/wiki/Photosynthesis)

Figure 7.2.7 shows the mean values of methane by all months between January and August. XCH4 is quite steady while seasons pass but there is a slightly decrease in March and an increase of about 2 % in the middle of July until the first days of August. However , in March and April due to bad weather that we had these days in Thessaloniki data are not so reliable as those in May , June , July or in February for example. Nevertheless , we can observe that the total column for XCH4 as it measures through the atmosphere with a portable FTIR spectrometer ( EM27/SUN ) is near stable during the most days with an exception around mid-February and in June 21th where abrupt changes are being shown.

XCH4 mean range between 1.79 ppm minimum and 1.84 ppm maximum along the seasons with very small standard deviation ( .

124

These differences state possibly due to the many different variable sinks and sources of methane, e.g. Dlugokencky et al. (1997).

A spread of a vortex of the wind could possibly lead to significantly altered CH4 values ( ppm ) . Atmospheric methane increased the past few years and has become the second largest contributor in radiative forcing although now it seems that have reached a steady state , with an overall atmospheric lifetime of about 9 years reflecting a balance between diverse sources (wetlands, ruminants, energy, rice agriculture, landfills, wastewater, biomass burning, oceans, and termites).

Figure 7.2.7 : XCH4 mean daily values in respect with dates - Thessaloniki

In the next figures we appose some examples of daily XCH4 measurements in Thessaloniki.

125

Figure 7.2.8 : XCH4 – HHMMSS in February 20th – Thessaloniki

Figure 7.2.9 : XCH4 – HHMMSS in June 21th – Thessaloniki

126

Figure 7.2.10 : XCH4 – HHMMSS in July 22th – Thessaloniki

Furthermore , The portable FTIR (Fourier transform infrared) spectrometer EM27/SUN, dedicated to the precise and accurate observation of column-averaged abundances of methane and carbon dioxide, has been equipped with a second detector channel, which allows the detection of additional species, especially carbon monoxide and the second band of methane. This allows an improved characterisation of observed carbon dioxide enhancements and makes the extended spectrometer especially suitable as a validation tool of ESA’s Sentinel 5 Precursor mission ( see chapter 8 ) , as it now covers the same spectral region as used by the infrared channel of the TROPOMI (TROPOspheric Monitoring Instrument) sensor. ( Frank Hase et al. , Addition of a channel for XCO observations to a portable FTIR spectrometer for greenhouse gas measurements ) .

127

The results of the retrievals in the 4208.7 – 4318 cm^−1 range is being compared to the reference values obtained from 5897 – 6145 cm^−1 (for CH4) as shown at the next figures.

Figure 7.2.11 : XCH4 daily mean values in respect with dates for the spectral region (4208.7,4318.8) – the same as the TROPOMI band covers

Juxtaposing these two different spectral bands for x- methane we can see some differences in the total column of XCH4. The TROPOMI band spectral coverage’s values are obviously larger than (5897.00,6145.00) band as we present in the following graph with a line 1:1 being concluded and a correlation factor calculated.

128

Figure 7.2.12 : Two different bands for XCH4 vs. date. The pattern is the same but obviously the concentrations are not.

129

Figure 7.2.13 : Validation between two different spectral regions for methane by FTIR ground-based measurements in Thessaloniki

In conclusion , we can observe that the correlation factor is quite close to the unity , so it means that they are in a good agreement. Also , XCH4_S5P_MEAN band overestimates in respect with the other band.

In the next step XCO , XH2O and XO2 daily mean values are presented from January 2019 until August 2019 for the region of the city of Thessaloniki.

Figure 7.2.14 : XCO daily mean values in respect with dates - Thessaloniki

130

Figure 7.2.15 : XH2O daily mean values in respect with dates - Thessaloniki

131

Figure 7.2.16 : XO2 daily mean values in respect with dates – Thessaloniki

Also , XCO mean values show a near steady course during the measurement days with a range between 0.08 ppm and 0.12 ppm. The primary source of carbon monoxide emissions is burning fossil fuels , fuel combustion and industrial processes.

Concerning that Thessaloniki is a big city ( second bigger city in Greece ) and FTIR ground-based instrument is placed in the city center of the city up on the roof of Aristotle University of Thessaloniki in the Laboratory of Atmospheric Physics close to the traffic area it is a natural consequence that CO emissions will be increased and also be as steady as it seems during the day ( but also the measurements days ).

As a consequence , we can conclude that not much to our surprise the highest concentrations of carbon monoxide tend to occur close to areas of high human population. On a global scale, this has meant that the more densely populated northern hemisphere has higher concentrations of carbon monoxide than the southern hemisphere. Biomass burning and fossil fuel use are the main sources of man-made carbon monoxide emissions.

XH2O and XO2 is being increased along the measurement period with the highest concentrations appear in June , July and August with a range between 200 and 7000 ppm.

132

Figure 7.2.17 : XCO – HHMMSS in July 18th – Thessaloniki

Figure 7.2.18 : XCO – HHMMSS in July 02th – Thessaloniki

133

Figure 7.2.19 : XH2O – HHMMSS in July 18th – Thessaloniki

134

Figure 7.2.20 : XO2 – HHMMSS in July 18th – Thessaloniki

7.3 ) Standard deviation of x-gases ( Daily values )

STANDARD DEVIATION OF X-GASES PER DAY Dates xco2_ xco_std xh2o_std xo2_std xch4_std (YYMM std DD) 190115 0.23754 0.00067 13.85057 10.69397 0.00047 190214 0.33657 0.00256 84.57308 73.44554 0.00081 190218 0.66770 0.00228 22.11728 12.51270 0.00117 190219 0.28395 0.00122 95.57930 85.63207 0.00126 190220 0.47265 0.00403 147.75536 108.37979 0.00272 190221 0.31948 0.00164 154.03738 119.03852 0.00231 190318 0.14746 0.00082 52.99029 33.76309 0.00179 190329 0.11286 0.00143 31.27669 23.86956 0.00129 190417 0.30330 0.00252 229.25077 182.36084 0.00116 190506 0.13344 0.00199 273.17345 229.57803 0.00066 190508 0.28011 0.00121 78.15116 72.29129 0.00365 190512 0.28609 0.00122 77.15553 61.05682 0.00113 190521 0.25955 0.00104 116.37175 102.70526 0.00182 190523 0.24143 0.00216 79.13816 55.79400 0.00266 190528 0.24144 0.00389 84.37423 62.03642 0.00351 190529 0.17552 0.00239 233.35506 174.60862 0.00273 190603 0.27803 0.00077 121.38974 97.30167 0.00084 190604 0.30110 0.00218 88.39516 77.02806 0.00192 190607 0.27293 0.00110 150.34580 146.37547 0.00113 190611 0.28929 0.00250 149.54516 112.14083 0.00284 190613 0.29226 0.00122 126.53684 86.68343 0.00229 190614 0.40294 0.00130 400.35790 348.27385 0.00285 190620 0.14831 0.00180 66.32886 46.10125 0.00152 190621 0.29031 0.00120 130.15022 101.59819 0.00457 190627 0.42739 0.00149 265.10421 222.19427 0.00244 190628 0.41820 0.00374 774.55626 616.27433 0.00108

135

190701 0.64836 0.00156 78.53940 62.56617 0.00467 190702 0.28349 0.00110 194.28226 169.15124 0.00344 190703 0.36293 0.00270 219.63885 178.86360 0.00279 190704 0.16530 0.00069 77.89190 57.83559 0.00124 190705 0.16260 0.00118 138.12308 103.92433 0.00243 190709 0.38886 0.00248 430.91590 358.95253 0.00111 190710 0.25680 0.00195 312.10764 251.57648 0.00082 190712 0.24861 0.00236 105.83947 89.12077 0.00474 190715 0.35149 0.00186 133.99056 115.26907 0.00107 190718 0.33940 0.00264 203.98536 138.20593 0.00126 190719 0.20862 0.00086 95.27122 72.04198 0.00201 190722 0.46718 0.00366 200.67562 177.93283 0.00372 190723 0.45458 0.00364 470.04391 349.07136 0.00077 190724 0.15137 0.00098 66.15785 50.27709 0.00084 190725 0.09589 0.00038 78.35172 58.43929 0.00068 190726 0.21614 0.00229 298.19566 205.20625 0.00109 190729 0.24090 0.00259 188.07735 143.57691 0.00150 190730 0.39831 0.00163 162.76401 107.58798 0.00085 190731 0.40663 0.00122 284.53958 190.75989 0.00207 190801 0.33448 0.00130 172.37115 146.73399 0.00243 Table 7.3 : Standard deviation of X-gases

Table 7.3 shows the standard deviation for all measured x-gases of each day and in this way we can observe by the following graphs how much close is every value to the mean one , so we can get a first impression of how closely the values converge each other.

In the next step of this work we quote all the figures that show the standard deviation of each day in respect always with the date of the year 2019 , where we started our measurements with EM27/SUN FTIR instrument , provided by Karlsruhe Institute of Technology ( KIT ).

136

Figure 7.3.1 : Standard deviation of XCO2 mean values of total column

Figure 7.3.2 : Standard deviation of XCH4 mean values of total column

137

Figure 7.3.3 : Standard deviation of XCH4_S5P2 mean values of total column

Figure 7.3.4 : Standard deviation of XCO mean values of total column

138

Figure 7.3.5 : Standard deviation of XH2O mean values of total column

Figure 7.3.6 : Standard deviation of XO2 mean values of total column

139

7.4 ) Comparison with other sites

Now lets see if our results in Thessaloniki are in good agreement with other European TCCON sites such as Bremen , Garmisch , Karlsruhe , Orleans for XCO2 for the time period 100301 – 120901 { YYMMDD }.

Sites Spring( ppm) Summer(ppm) Autumn(ppm) Winter(ppm) Thessaloniki 405-408.4 400.5-407.25 - 401-406.4 Bremen 389.08-394.1 388.2-392.1 383-385 389-396 Garmisch 390.24-392.5 386-390 383.9-387 389.5-392.6 Karlsruhe 390.5-393.8 385.9-389.95 383-388.7 392-396.85 Orleans 390-394.2 383.11-388.95 - 392-395.48 Table 7.4.1 : Daily mean FTIR XCO2 data of four European TCCON sites: Bremen, Garmisch , Karlsruhe and Orleans for the year 2010 and Thessaloniki ( our own measurements ) for the year 2019 comparison (Germany, Messerschmidt et al. [2011])

Next , total columns of XCO2 and XCH4 daily course in Paris Campaign in 2015 are compared with Thessaloniki FTIR ground – based EM27/SUN instrument results in the beginning of May of 5 different .

Also , Berlin Campaign is taking into consideration to show the differences with our site in Thessaloniki , Greece with 5 different sites where the distances between the sites are about 25 km or less .

140

PARIS CAMPAIGN

Figure 7.4.1 : XCO2 daily course at 2015/05/07

Figure 7.4.2 : XCH4 daily course at 2015/05/07

141

(MIT, GIF, PIS, RES, JUS) are the five different sites from the top to the bottom Source : XCO2 in an emission hot-spot region: the COCCON Paris campaign 2015 Felix R. Vogel, Matthias Frey , Johannes Staufer , Frank Hase , Grégoire Broquet, Irène Xueref-Remy , Frédéric Chevallier , Philippe Ciais , Mahesh Kumar Sha, Pascale Chelin , Pascal Jeseck , Christof Janssen , Yao Té , Jochen Groß , Thomas Blumenstock , Qiansi Tu, and Johannes Orphal

Figure 7.4.3 : XCO2 daily course at 2019/05/08 in Thessaloniki

142

Figure 7.4.4 : XCH4 daily course at 2019/05/08 in Thessaloniki

We can conclude that in Thessaloniki the total column from XCO2 in higher than the other 5 sites , but XCH4 is quite same at those 6 different locations in Europe.

In Thessaloniki due to the large number of vehicles and human population in the city center is quite normal that it has the highest concentrations of XCO2.

143

BERLIN CAMPAIGN

Figure 7.4.5 : XH2O (top panel), CXO2 (middle panel), and XCH4 (bottom panel) as measured at all sites during the Berlin campaign , in July and August 2014. Source : Application of portable FTIR spectrometers for detecting greenhouse gas emissions of the major city Berlin F. Hase, M. Frey, T. Blumenstock, J. Groß, M. Kiel, R. Kohlhepp, G. Mengistu Tsidu, K. Schäfer, M. K. Sha, and J. Orphal

144

Figure 7.4.6 : Observed daily variability of carbon dioxide (XCO2) and methane (XCH4) during 27 June 2014 Source : Application of portable FTIR spectrometers for detecting greenhouse gas emissions of the major city Berlin F. Hase, M. Frey, T. Blumenstock, J. Groß, M. Kiel, R. Kohlhepp, G. Mengistu Tsidu, K. Schäfer, M. K. Sha, and J. Orphal

145

Locations XCO2 XCH4 Thessaloniki 402.2-404.3 1.812-1.825 Mahlsdorf 395.9-393.25 1.806-1.815 Charlottenberg 397.5-392.9 1.809-1.816 Heiligensee 399.2-392.4 1.808-1.822 Lindenberg 397.12-394.24 1.805-1.814 Lichtenarde 395.84-393.15 1.806-1.824 Table 7.4.2 : Comparison of XCO2 and XCH4 between Thessaloniki (2019) in Greece and 5 different sites near Berlin in Berlin campaign in 2015 during 27/06 in summer.

Figure 7.4.7 : XCO2 daily variability during 27 June 2019 in Thessaloniki

146

Figure 7.4.8 : XCH4 daily variability during 27 June 2019 in Thessaloniki

Juxtaposing these two different countries we can observe that in Thessaloniki the total columns of carbon dioxide and methane are larger than the five different sites around Berlin. This results from the fact that Thessaloniki is a bigger city than these 5 near Berlin , also the human population is a significant contributor in the city center of Thessaloniki , thereby XCO2 concentrations should be increased in respect with the other five of Germany. Concerning methane total column we can see that they are in a pretty good agreement in June 26th but we must point out the fact that the years of measurement are different { 2014, 2019 }. Southerly winds prevailed during that day, and indeed the XCO2 values observed in Heiligensee in the northwest of Berlin are elevated. It is important to note that, although the emission signals tend to be smaller than the observed intraday variability, enhancements as small as 0.5‰ are noticeable. In detail, the observed XCH4 enhancements differ from the XCO2 enhancements, which is expected due to different sources. Moreover, the background of the XCH4 seems less well defined and more variable. This meets the expectation: due to the likely presence of rural CH4 sources around the conurbation area encircled with the stations and due to the stronger contrast between tropospheric and stratospheric

147 mixing ratios of CH4, higher variability is expected in the XCH4 background field than in case of XCO2. (Source : Application of portable FTIR spectrometers for detecting greenhouse gas emissions of the major city Berlin F. Hase, M. Frey, T. Blumenstock, J. Groß, M. Kiel, R. Kohlhepp, G. Mengistu Tsidu, K. Schäfer, M. K. Sha, and J. Orphal )

8) Validation with TROPOMI satellite sensor ( Sentinel - S5P )

Introduction

In situ measurements , satellite observations and ground- based remote sensing measurements

Trace gases account for less than 1 % of the Earth’s atmospheric chemical composition but strongly influence the climate system. The increase of trace gas concentrations has a large impact on the Earth’s radiative budget and substantially enhances the greenhouse effect. The radiative forcing of well mixed greenhouse gases (GHG) increased by 2.83 Wm^(−2) from the start of the industrial era in the year 1750 to 2011 (IPCC, 2013). This change is largely due to the increase of carbon dioxide (CO2) concentrations in the atmosphere. Over the last decade, the radiative forcing of CO2 has a growth rate of 0.27 Wm^(−2) per decade. The net radiative forcing of GHGs other than CO2 accounted for about 1.00 Wm^(−2) between 1750 and 2011 (IPCC, 2013). The growth in concentrations of GHG’s and other trace gases can be largely attributed to anthropogenic emissions, mainly driven by energy related industries burning fossil fuels and land use change (Le Quιrι et al., 2015). This leads to a continuous increase of the radiative forcing affecting the global climate. In the recent years, great effort was undertaken to improve the understanding of the changing climate. Several

148 measurement programs were initiated to monitor trace gas concentrations and long-term trends in the Earth’s atmosphere. These programs include observations of the atmosphere and surface, aiming at providing long-term regional and global climate quality data records to support scientific studies and future climate projections (e.g. Meul et al.,2016). The assimilation of different measurement methods, including in situ measurements, satellite observations and ground-based remote sensing measurements provide a unique data set which yields results that could not have been uncovered by one single measurement method. Ground-based remote sensing measurements provide a transfer standard between the highly precise but locally limited in situ measurements and global observations by satellite-borne instruments which are limited in precision and temporal resolution. An established ground-based remote sensing technique is provided by the Fourier Transform Infrared (FTIR) Spectroscopy using direct Sun light as source to measure gas abundances in the Earth’s atmosphere. FTIR instruments measure quantities comparable to satellite data but are more precise than space-borne observations , due to the spatial coverage which they obtain. ( source : The remote sensing of tropospheric composition from space, Editors John Burrows , Ulrich Platt , Peter Borrel )

149

8.1) Sentinel 5 – Precursor

Figure 8.1 : Sentinel-5P carries the state-of-the-art Tropomi instrument, which maps a multitude of trace gases such as nitrogen dioxide, ozone, formaldehyde, sulphur dioxide, methane, carbon monoxide and aerosols. (Image credit ESA/ATG medialab)-( https://physicsworld.com/a/tropomi-finds-source-of-pollutants/)

The Sentinels are a fleet of satellites designed specifically to deliver the wealth of data and imagery that are central to the European Commission’s Copernicus programme. This unique environmental monitoring programme is making a step change in the way we manage our environment, understand and tackle the effects of climate change and safeguard everyday lives.

Sentinel-5 Precursor – also known as Sentinel-5P – is the first Copernicus mission dedicated to monitoring our atmosphere. It observes sunlight that is scattered back to space by Earth’s surface and atmosphere, detecting the unique fingerprints of gases in different parts of the spectrum.

150

The satellite carries the state-of-the-art Tropomi instrument to map a multitude of trace gases such as nitrogen dioxide, ozone, formaldehyde, sulphur dioxide, methane, carbon monoxide and aerosols – all of which affect the air we breathe and therefore our health, and our climate. What sets Tropomi apart is that it measures in the ultraviolet and visible (270–500 nm), near-infrared (675–775 nm) and shortwave infrared (2305–2385 nm) spectral bands. This means that a wide range of pollutants such as nitrogen dioxide, ozone, formaldehyde, sulphur dioxide, methane and carbon monoxide can be imaged more accurately than ever before. With a swath width of 2600 km, it will map the entire planet every day. Information from this new mission will be used through the Copernicus Atmosphere Monitoring Service for air quality forecasts and for decision-making. The TROPOspheric Monitoring Instrument (TROPOMI) is the satellite instrument on board the Copernicus Sentinel-5 Precursor satellite. The Sentinel-5 Precursor (S5P) is the first of the atmospheric composition Sentinels, launched on 13 October 2017, planned for a mission of seven years.

Tropomi will map the global atmosphere every day with a resolution as high as 7 km × 3.5 km. At this resolution, air pollution over cities can be detected. The instrument, the single payload of the Sentinel-5P spacecraft, uses passive remote sensing techniques to attain its objective by measuring, at the Top Of Atmosphere (TOA), the solar radiation reflected by and radiated from the earth. The mission also contributes to services such as volcanic ash monitoring for aviation safety and for services that warn of high levels of UV radiation which can cause skin damage. In addition, scientists will also use the data to improve our knowledge of important processes in the atmosphere related to the climate and to the formation of holes in the ozone layer. Sentinel-5P is the result of close collaboration between ESA, the European Commission, the Netherlands Space Office, industry, data users and scientists. The mission has been designed and built by a consortium of 30 companies led by Airbus Defense and Space UK and NL. Sentinel-5P was developed to reduce data gaps between the Envisat satellite – in particular the Sciamachy instrument – and the launch of Sentinel-5, and to complement GOME-2 on MetOp.

( source https://sentinel.esa.int/web/sentinel/user-guides/sentinel-5p-tropomi )

151

Figure 8.2 : Sentinel -5P- TROPOMI ( source https://sentinel.esa.int/web/sentinel/user- guides/sentinel-5p-tropomi )

In the future, both the geostationary Sentinel-4 and polar-orbiting Sentinel-5 missions will monitor the composition of the atmosphere for Copernicus Atmosphere Services. Both missions will be carried on meteorological satellites operated by Eumetsat. Until then, the Sentinel- 5P mission will play a key role in monitoring and tracking air pollution. https://www.esa.int/Our_Activities/Observing_the_Earth/Copernicus/Sentinel- 5P/Data_products

152

Figure 8.3 : Data products from Sentinel-5P’s Tropomi instrument that are distributed to users at two levels: Source(https://www.esa.int/Our_Activities/Observing_the_Earth/Copernicus/Sentinel -5P/Data_products)

The levels that we mentioned above are the following :

Level-1B: provides geo-located and radiometrically corrected top of the atmosphere Earth radiances in all spectral bands, as well as solar irradiances.

Level-2: provides atmospheric geophysical parameters.

The various applications identified for Sentinel-5P require a systematic sampling of a set of atmospheric components both at global and regional scales, with focus at the European area. In line with the needs of numerical forecasting systems, Level-2 products are to be supplied within three hours after sensing. This 'near real-time' (NRT) service delivers products for ozone, sulphur dioxide, nitrogen dioxide, formaldehyde and carbon monoxide, vertical profiles of ozone as well as cloud/aerosol distributions.

Source https://www.esa.int/Our_Activities/Observing_the_Earth/Copernicus/Sentinel- 5P/Data_products

To sum up the main characteristics of TROPOMI are:

153

 Type: passive grating imaging spectrometer  Configuration: Push broom staring (non-scanning) in nadir viewing  Swath width: 2,600 km 2  Spatial sampling: 7x7 km  Spectral: 4 spectrometers, each electronically split in two bands (2 in UV, 2 in VIS, 2 in NIR, 2 in SWIR)  Radiometric accuracy (absolute): 1.6% (SWIR) to 1.9% (UV) of the measured earth spectral reflectance.  Overall mass: 204.3 kg not including ICU (16.7 kg) that is integrated on the platform, separated from the instrument.  Dimensions: 1.40 x 0.65 x 0.75 m  Design lifetime: 7 years  Average power consumption: 155 W  Generated data volume: 139 Gbits per full orbit.

Figure 8.4 : TROPOMI instrument (credit: TNO, Airbus DS-NL)- (https://sentinels.copernicus.eu/web/sentinel/missions/sentinel-5p/instrumental-payload)

154

8.2) EM27/SUN validation with TROPOMI data products (XCO,XCH4)

In this study we will present the comparison between FTIR ground-based measurements of total column averaged dry-air mole fractions ( XCO and XCH4 ) - taken in the city center of Thessaloniki upon the roof of Aristotle University of Thessaloniki in the Laboratory of Atmospheric Physics with a portable spectrometer , EM27/SUN - and space-based retrievals of Sentinel S5P which was launched in October 13th of 2017.

In the following figures we will see if FTIR measurements fall well within satellite ( TROPOMI , SENTINEL S5P ) observations. In a first step we will present time series from Sentinel S5P and FTIR ground – based measurements with common days of measurements and the time series with separate days , because the days that the satellite was able to take precise values of these x-gases were not exactly the same as the number of FTIR measurements which were taken in Thessaloniki. Also we will appose 1:1 line to see how close are the FTIR measurements with the spaced-observed one.

So , it is important to clarify that we select only the values of TROPOMI data which were not so cloudy with a q.a. value above 40 % so that the cloud fraction would not affect the scanning of satellite to much. Furthermore , in order to keep reliable measurements we tried to isolate only data as close as we could from the coordinates of Thessaloniki. ( We mentioned before that TROPOMI covers a long distance of measure with many CCD’s ) . However , for example methane has a long lifetime in the atmosphere so it is not so primary to obtain only these values , because of the fact that it is well-mixed in the atmosphere.

In the following figures we can observe the differences between satellite data and FTIR ground – based measurements and how much one approaches other.

155

Figure 8.2.1 : Time series of methane with ground – based FTIR EM27/SUN and with TROPOMI satellite sensor of Sentinel S5P with common measurement-days from January 15th until July 1st

156

Figure 8.2.2 : Time series of methane with ground – based FTIR EM27/SUN and with TROPOMI satellite sensor of Sentinel S5P from January 1st until July 13th

157

Figure 8.2.3 : Scatter plot of FTIR MEASURED XCH4 and satellite- retrieved methane. The solid line is for 1:1 respectively.

We can conclude that the satellite overestimates in most cases, except from one day, with a range between 1.807 ppm and 1.882 ppm. We expected to see a better correlation between FTIR ground-based and satellite data because of the fact that methane is near steady among other gases in the atmosphere. It can be estimated that the deviations from satellite sensor , TROPOMI , may point out the big coverage of Tropomi which map the global atmosphere every day with a resolution as high as 7 km × 3.5 km. Also , there are uncertainties due to the fact that a ground-based FTIR instrument taking spectra during the whole day that the sun is clear of clouds , but on the other hand a satellite sensor monitors the whole planet and take measurements only at 13:30 UTC every day in case that the cloud fraction is not so high in order to make our observations more precise. We hope that in a longer period of ground-based FTIR measurements the correlation factor would be better and closely to the unity , supposed that the larger number of total measurements will eliminate these deviations from our own measurements.

Finally , it can be seen that methane is quite stable during the changes of periods and range between 1.775 ppm and 1.882 ppm possible from agriculture. We observe that the days that we see an increase in methane total column were windy days so the increase may be due to the transportation of wind from the areas near Thessaloniki , where there are fields and livestock farming , a primary source of methane concentrations as well.

For example in February - the wind was strong the whole day , so it can be clearly estimated that the increase from methane was due to the transportation of air which “brought” methane concentrations in the city center of Thessaloniki.

In the following figure we see a plot of methane vs. time for the region of Thessaloniki in February 20th , in which can be observed a slightly increase in midday ( always the time is in UTC ).

158

Figure 8.2.4 : Daily XCH4 with time in Thessaloniki in the winter In the next step we are going to present the correlation between FTIR and TROPOMI data product carbon monoxide ( XCO ) .

159

Figure 8.2.5 : Time series of carbon monoxide with ground – based FTIR EM27/SUN and with TROPOMI satellite sensor of Sentinel S5P with common measurement-days from January 15th until July 1st

160

Figure 8.2.6 : Time series of carbon monoxide with ground – based FTIR EM27/SUN and with TROPOMI satellite sensor of Sentinel S5P from January 1st until July 13th

161

Figure 8.2.7 : Scatter plot of FTIR MEASURED XCO and satellite- retrieved carbon monoxide. The solid line is for 1:1 respectively.

In conclusion , there is a slightly increase of carbon monoxide in spring but in general is more and less near steady along days of measurement. It is of primary importance to notice that FTIR ground- based instrument ( EM27/ SUN BY BRUKER OPTICS ) overestimates in this case and the satellite sensor TROPOMI underestimates carbon monoxide measurements( Figure 8.2.7). XCO values range between 0.06 ppm and 0.12 ppm with a correlation factor of R^2 = 0.448 which is by far better correlation than XCH4 ( methane ) that we showed in our work later in the previous graphs.

The highest concentrations of carbon monoxide tend to occur close to areas of high human population and Thessaloniki is one of these areas. This means that the primary source of carbon monoxide is by fuel combustion and transportation with burning fossil fuels , on road vehicles and engines. Almost 60 % of carbon monoxide sources come from vehicles and fossil fuels or engines , but also another source is the industrialization of bnig cities. ( epa.gov/greenhousegases/carbon monoxide )

162

Figure 8.2.8 : Daily XCO with time in Thessaloniki in the winter

Figure 8.2.9 : Daily XCO with time in Thessaloniki in the summer

In above graphs we can see a daily course of carbon monoxide in the winter but also in the summer where we can observe how steady is in the atmosphere except from midday in 190220 ( figure 8.2.8 )

163

Finally we can conclude that the differences between ground based and satellite data are due to the spatial coverage of TROPOMI , because CO is increasing due to human transport by vehicles which takes place in the heart of the city of Thessaloniki , where the FTIR EM27/SUN is placed. So , it is logical that the higher concentrations should be in the city center where all the traffic occurs.

9) Conclusions

In our days, global climate change is one of the most urgent challenges facing man-kind, however it is induced by humans themselves. Industrialization and land use change cause a significant modification of the atmospheric composition leading to major implications like the depletion of stratospheric ozone or global warming. The latter one is continuously boosted by increasing greenhouse gas (GHG) concentrations.

In order to understand global climate change and to estimate future effects, the investigation of CO2 sources and sinks within the carbon cycle is strongly required. For this reason, several measurement programs were initialized being mainly in situ measurements, satellite measurements and ground-based Fourier transform infrared (FTIR) measurements. In contrast to in situ measurements, ground-based FTIR spectrometers and satellites detect total column abundances of CO2.

Furthermore it is of primary importance to validate ground-based FTIR measurements with satellite data products in order to estimate as good as we can sources , sinks or emissions from these controversial X-gases.

Although the greenhouse effect is a natural process where the atmosphere traps some of the sun's energy, warming the Earth enough to support life, humans have greatly increased the concentrations of greenhouse gases, thus causing a significant increase in the overall greenhouse effect. A number of gases are involved in the human caused

164 enhancement of the greenhouse effect. These gases include: carbon dioxide (CO2); methane (CH4); nitrous oxide (N20); CFC's and ozone (03) Out of all these gases the most important is carbon dioxide which accounts for around 55% of the change in the intensity of the Earth's greenhouse effect. The consequence of the greenhouse effect is that there will a rise in the sea levels around the world, there will be dramatic climate changes, and agriculture will suffer from the fluxes of the weather.

However, it's not too late to cut back on greenhouse gas emissions and some effective ways to reduce emissions is to use cleaner fuels, use energy efficient machines, develop alternative sources for energy and to plant more trees

165

References :  Aeronomy of the Middle Atmosphere Chemistry and Physics of the Stratosphere and Mesosphere -- Third revised and enlarged edition Guy P.  An Introduction to Atmospheric Radiation , Second Edition K. N. Liou -- DEPARTMENT OF ATMOSPHERIC SCIENCES UNIVERSITY OF CALIFORNIA , LOS ANGELES, CALIFORNIA Amsterdam Boston  A.G. Marshall, Acc. Chem. Res., 1985, 18, 316  R. Korb, P. Dybwad, W. Wadsworth, J. W. Salisbury, Applied Optics, 1996, 35, 1679  Ahonen, H. Riipinen, A. Roos, Analyst, 1996, 121, 1253  L. Smith, Applied Infrared Spectroscopy : Fundamentals, Techniques, and Analytical Problem-solving, Wiley, New York, 1979  Assessment of errors and biases in retrievals of XCO2, XCH4, XCO, and XN2O from a 0.5 cm^-1 resolution solar-viewing spectrometer (Atmos. Meas. Tech., 9, 3527–3546, 2016 www.atmos-meas-tech.net/9/3527/2016/ doi:10.5194/amt-9-3527-2016 © Author(s) 2016. CC Attribution 3.0 License. )  C. Smith, Fundamentals of Fourier Transform Infrared Spectroscopy, CRC press, 1996  Stuart, Modern Infrared Spectroscopy, Wiley, New York, 1996  Bernhard Schrader(editor): Infrared and Raman spectroscopy, VCH, Weinheim, Federal Republic of Germany , 1995  Bernhard Jaun, Analytische Chemie IV-Structure determination by NMR, 1. Practical aspects of pulse Fourier transform NMR spectroscopy, 1.1-1.21  Chen, J., Viatte, C., Hedelius, J. K., Jones, T., Franklin, J. E., Parker, H., Gottlieb, E. W., Wennberg, P. O., Dubey, M. K., and Wofsy, S. C.: Differential column measurements using compact solar-tracking spectrometers, Atmospheric Chemistry and Physics, 16, 8479–8498, https://doi.org/10.5194/acp-16-8479- 2016, https://www.atmos-chem-phys.net/16/8479/2016/, 2016.  Calibration and instrumental line shape characterization of a set of portable FTIR spectrometers for detecting greenhouse gas emissions (Karlsruhe Institute of Technology (KIT), Institute of Meteorology and Climate Research (IMK-ASF), Karlsruhe, Germany ) Atmos. Meas. Tech., 8, 3047–3057, 2015 www.atmos-meas-tech.net/8/3047/2015/ doi:10.5194/amt-8-3047-2015 © Author(s) 2015. CC Attribution 3.0 License.  Camtracker: a new camera controlled high precision solar tracker system for FTIR-spectrometers M. Gisi, F. Hase, S. Dohe, and T. Blumenstock (Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climate Research (IMK-ASF), Karlsruhe, Germany) Atmos. Meas. Tech., 4, 47–54, 2011 www.atmos-meas-tech.net/4/47/2011/ doi:10.5194/amt-4-47-2011 © Author(s) 2011. CC Attribution 3.0 License.

166

 Chen, Y.-H., and R.G. Prinn, 2006: Estimation of atmospheric methane emissions between 1996-2001 using a 3D global chemical transport model. J. Geophys. Res., 111, D10307, doi:10.1029/2005JD006058.  Center for Sustainable Systems, University of Michigan. 2018. “Climate Change: Science and Impacts Factsheet.” Pub. No. CSS05-19.  CSS calculation based on data from UNEP and UN Framework Convention on Climate Change (UNFCCC) (2003) Climate Change Information Kit.  Davis, S., Abrams, M. C., and Brault, J. W.: Fourier Transform Spectrometry, Academic press, 2010. Dlugokencky, E. J., Masarie, K. A., Tans, P. P., Conway, T. J., and Xiong, X.: Is the amplitude of the methane seasonal cycle changing?, Atmospheric Environment, 31, 21–26, https://doi.org/10.1016/S1352- 2310(96)00174 4, http://www.sciencedirect.com/science/article/pii/ S1352231096001744, 1997. Dohe, S.: Measurements of atmospheric CO2 columns using ground-based FTIR spectra, Ph.D. thesis, Karlsruhe Institute of Technology, 2013.  D.A. Skoog and J.J. Leary. “Principles of Instrumental Analysis, 4th Ed.”, Harcourt Brace Jovanovich. Philadelphia, PA, 1992. Chapter 12.  Daniels, J.W. Williams, P. Bender, R.A. Alberty, C.D. Cornwell, J. E. Harriman. "Experimental Physical Chemistry, 7th Ed.”, McGraw-Hill, New York, NY, 1970.  W. Ball, Field Guide to Spectroscopy, SPIE Publication, Bellingham, 2006  Dentener, F., et al., 2006: Emissions of primary aerosol and precursor gases in the years 2000 and 1750 - prescribed data-sets for AeroCom. Atmos. Chem. Phys. Discuss., 6, 2703–2763.  Department of Low and Medium Energy Physics-F2, Josef Stefan Institute, Ljubljana, Slovenia  Donald A. Burns, Emil W. Ciurczak: Handbook of Near-Infrared Analysis, second edition, Marcel Dekker Inc., New York-Basel, 2001

 Frey ,M., Hase, F., Blumenstock, T., Groß, J., Kiel,M.,Mengistu Tsidu, G., Schäfer, K., Sha,M. K., and Orphal, J.: Calibration and instrumental line shape characterization of a set of portable FTIR spectrometers for detecting greenhouse gas emissions, Atmospheric Measurement Techniques, 8, 3047– 3057, https://doi.org/10.5194/amt-8-3047-2015, https://www.atmos-meas- tech.net/8/3047/2015/, 2015.  Hase, F., Frey, M., Kiel, M., Blumenstock, T., Harig, R., Keens, A., and Orphal, J.: Addition of a channel for XCO observations to a portable FTIR spectrometer for greenhouse gas measurements, Atmospheric Measurement Techniques, 9, 2303{2313, https://doi.org/10.5194/amt-9- 2303-2016, 2016.

167

 Hase, F., Blumenstock, T., and Paton-Walsh, C.: Analysis of the instrumental line shape of highresolution  Fourier transform IR spectrometers with gas cell measurements and new retrieval software, Appl. Opt., 38, 3417{3422, https://doi.org/10.1364/AO.38.003417, 1999.  Hase, F., Hannigan, J., T. Co_ey, M., Goldman, A., H•opfner, M., Jones, N., P. Rinsland, C., and Wood, S.: Intercomparison of retrieval codes used for the analysis of high-resolution, groundbased FTIR measurements, Journal of Quantitative Spectroscopy & Radiative Transfer, 87, 25{52, https://doi.org/10.1016/j.jqsrt.2003.12.008, 2004.  Hase, F., Drouin, B. J., Roehl, C. M., Toon, G. C., Wennberg, P. O., Wunch, D., Blumenstock, T., Desmet, F., Feist, D. G., Heikkinen, P., De Mazi_ere, M., Rettinger, M., Robinson, J., Schneider, M., Sherlock, V., Sussmann, R., T_e, Y., Warneke, T., and Weinzierl, C.: Calibration of sealed HCl cells used for TCCON instrumental line shape monitoring, Atmospheric Measurement Techniques, 6, 3527{3537, https://doi.org/10.5194/amt-6-3527-2013, 2013.  Fourier Transform Infrared Spectrometry, 2nd Edition Peter R. Griffiths, James A. De Haseth, James D. Winefordner (Series Editor)  Hase et al.: “The portable EM27/SUN FTIR spectrometer for GHG measurements”  Fundamentals of Fourier Transform Infrared spectroscopy @ Brian. C. Smith  Global Warming – The complete briefing ( Fourth Edition )-John Houghton(Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK)

 Gisi, M., Hase, F., Dohe, S., and Blumenstock, T.: Camtracker: a new camera controlled high precision solar tracker system for FTIRspectrometers, Atmospheric Measurement Techniques, 4, 47–54, https://doi.org/10.5194/amt-4-47-2011, https://www.atmos-meas-tech. net/4/47/2011/, 2011.  Gisi, M., Hase, F., Dohe, S., Blumenstock, T., Simon, A., and Keens, A.: XCO2- measurements with a tabletop FTS using solar absorption spectroscopy, Atmospheric Measurement Techniques, 5, 2969–2980, https://doi.org/10.5194/amt-5-2969-2012, http://www. atmos-meas-t Gri_ths, P. and de Haseth, J. A.: Fourier Transform Infrared Spectrometry, Chemical Analysis: A Series of Monographs on Analytical Chemistry and Its Applications, Wiley&Sons, New York, 2007.ech.net/5/2969/2012/, 2012.  Hase, F.: Improved instrumental line shape monitoring for the ground-based, high-resolution FTIR spectrometers of the Network for the Detection of Atmospheric Composition Change, Atmospheric Measurement Techniques, 5, 603–610, https://doi.org/10.5194/amt-5-603-2012, https://www.atmos- meas-tech.net/5/603/2012/, 2012.

168

 Hase, F., Hannigan, J., T. Coffey, M., Goldman, A., Höpfner, M., Jones, N., P. Rinsland, C., andWood, S.: Intercomparison of retrieval codes used for the analysis of high-resolution, ground-based FTIR measurements, Journal of Quantitative Spectroscopy & Radiative Transfer, 87, 25–52, 2004.  Hase, F., Frey, M., Blumenstock, T., Groß, J., Kiel, M., Kohlhepp, R., Mengistu Tsidu, G., Schäfer, K., Sha, M. K., and Orphal, J.: Application of portable FTIR spectrometers for detecting greenhouse gas emissions of the major city Berlin, Atmospheric Measurement Techniques, 8, 3059–3068, https://doi.org/10.5194/amt-8-3059-2015, https://www.atmos-meas- tech.net/8/3059/2015/, 2015.  Hase, F., Hannigan, J., T. Co_ey, M., Goldman, A., H•opfner, M., Jones, N., P. Rinsland, C., and Wood, S.: Intercomparison of retrieval codes used for the analysis of high-resolution, groundbased FTIR measurements, Journal of Quantitative Spectroscopy & Radiative Transfer, 87, 25{52, https://doi.org/10.1016/j.jqsrt.2003.12.008, 2004.  Hase, F., Drouin, B. J., Roehl, C. M., Toon, G. C., Wennberg, P. O., Wunch, D., Blumenstock, T., Desmet, F., Feist, D. G., Heikkinen, P., De Mazi_ere, M., Rettinger, M., Robinson, J., Schneider, M., Sherlock, V., Sussmann, R., T_e, Y., Warneke, T., and Weinzierl, C.: Calibration of sealed HCl cells used for TCCON instrumental line shape monitoring, Atmospheric Measurement Techniques, 6, 3527{3537, https://doi.org/10.5194/amt-6-3527-2013, 2013.  Hase, F., Blumenstock, T., and Paton-Walsh, C.: Analysis of the Instrumental Line Shape of High-Resolution Fourier Transform IR Spectrometers with Gas Cell Measurements and New Retrieval Software, Applied Optics, 38, 3417{ 3422, doi:10.1364/AO.38.003417, URL http://ao.osa.org/abstract.cfm?URI= ao-38-15-3417, 1999.  Hartmann E, et al. (1994)  Hassen Aroui, Johannes Orphal and Fridolin Kwabia Tchana (2012). Fourier Transform Infrared Spectroscopy for the Measurement of Spectral Line Profiles, Fourier Transform - Materials Analysis, Dr Salih Salih (Ed.), ISBN: 978- 953-51-0594-7, InTech, Available from: http://www.intechopen.com/books/fourier-transformmaterials- analysis/fourier-transform-infrared-spectroscopy-for-the-measurement-of- spectral-line-profiles  Humlicek, J.: An E_cient Method for Evaluation of the Complex Probability Function: The Voigt Function and its Derivatives, Journal of Quantitative Spectroscopy and Radiative Transfer, 21, 309{313, https://doi.org/10.1016/0022-4073(79)90062-1, 1979.  IPCC-WG1: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I for the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K.

169

Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)], Cambridge University Press, https://doi.org/ 10.1017/CBO9781107415324, 2013.  IPCC-WG3: Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y. Sokona, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J. Savolainen, S. Schlomer, C. von Stechow, T. Zwickel and J.C. Minx (eds.)], Cambridge University Press, https://doi.org/ 10.1017/CBO9781107415416, 2014.  IPCC: Climate Change 2001: Synthesis Report. A Contribution of Working Groups I, II and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge and , New York, 2001.  IPCC: Climate Change 2007: Synthesis Report. A Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, IPCC, Geneva, 2007.  IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp.  IPCC (2014) Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.  https://doi.org/10.1364/AO.46.004774, http://ao.osa.org/abstract. cfm?URI=ao-46-21-4774, 2007.  John Chalmers and Peter Gri_ths: Introduction to the theory and practise of vibrational spectroscopy, John WIley and Sons, 2002  Klappenbach, F., Bertleff, M., Kostinek, J., Hase, F., Blumenstock, T., Agusti- Panareda, A., Razinger, M., and Butz, A.: Accurate mobile remote sensing of XCO2 and XCH4 latitudinal transects from aboard a research vessel, Atmospheric Measurement Techniques, 8, 5023– 5038, https://doi.org/10.5194/amt-8-5023-2015, https://www.atmos-meas- tech.net/8/5023/2015/, 2015.  Keeling, C.D., and T.P. Whorf, 2005: Atmospheric CO2 records from sites in the SIO air sampling network. In: Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN, http://cdiac.esd.ornl.gov/trends/co2/sio-keel-fl ask/sio-keel-fl ask.html.

170

 Keppel-Aleks, G., Toon, G. C., Wennberg, P. O., and Deutscher, N. M.: Reducing the impact of source brightness uctuations on spectra obtained by Fourier-transform spectrometry, Appl. Opt., 46, 4774{4779, https://doi.org/10.1364/AO.46.004774, 2007.  Kiel, M., Hase, F., Blumenstock, T., and Kirner, O.: Comparison of XCO abundances from the Total Carbon Column Observing Network and the Network for the Detection of Atmospheric Composition Change measured in Karlsruhe, Atmospheric Measurement Techniques, 9, 2223{  2239, https://doi.org/10.5194/amt-9-2223-2016, 2016a.  Lars-Erik Amand and Claes J. Tullin: The theory behind FTIR analysis, written documentation prepared for a distance course given by The Centre of Combustion Science and Technology, CECOST on "Measurement Technology", Lund, Sweden  Marland G (2006) The human component of the carbon cycle. Testimony before the Committee on Government Reform, Subcommittee on Energy and Resources, 27 September 2006, US House of Representatives, Washington DC  Messerschmidt, J., Geibel, M. C., Blumenstock, T., Chen, H., Deutscher, N. M., Engel A., Feist, D. G., Gerbig, C., Gisi, M., Hase, F., Katrynski, K., Kolle, O., Lavri_c, J. V., Notholt, J., Palm, M., Ramonet, M., Rettinger, M., Schmidt, M., Sussmann, R., Toon, G. C., Truong, F., Warneke, T., Wennberg, P. O., Wunch, D., and Xueref- Remy, I.: Calibration of TCCON column-averaged CO2: the _rst aircraft campaign over European TCCON sites, Atmospheric Chemistry and Physics, 11, 10 765{ 10 777, doi:10.5194/acp-11-10765-2011, URL http://www.atmos-chem-phys.net/ 11/10765/2011/, 2011.  Olsen, S. C. and Randerson, J. T.: Differences between surface and column atmospheric CO2 and implications for carbon cycle research, Journal of Geophysical Research: Atmospheres, 109, https://doi.org/10.1029/2003JD003968,http://dx.doi.org/10.1029/2003JD00 3968, 2004.  P.R. Griffiths, Science, 21, 1983, 297  P. R. Griffiths, J. A. de Haseth, Fourier Transform Infrared Spectrometry, Wiley, New York, 1986  P. R. Griffiths, Chemical Infrared Fourier Transform Spectroscopy, Wiley, New York, 1975  Peter R. Griffiths and James A. de Haseth: Fourier Transform Infrared Spectrometry, Second edition, John WIley and Sons, Hoboken, New Jersey, 2007  Peter Atkins, Julio De Paula. 2006. Physical Chemistry, 8th ed. Oxford University Press: Oxford, UK.  Peter J. Larkin: IR and Raman spectroscopy: Principles and Spectral Interpretation, Elsevier Inc., 2011

171

 Platt, U. and J. Stutz, 2008, Differential optical absorption spectroscopy: principles and applications, Springer Verlag, Heidelberg, ISBN 978- 3540211938, pp597.  Proposal 20140246, Department of Low and Medium Energy Physics-F2, Josef Stefan Institute, Ljubljana, Slovenia  Rothman, L. S., Gordon, I. E., Barbe, A., Benner, D. C., Bernath, P. F., Birk, M., Boudon, V., Brown, L. R., Campargue, A., Champion, J.-P., Chance, K., Coudert, L. H., Dana, V., Devi, V. M., Fally, S., Flaud, J.-M., Gamache, R. R., Goldman, A., Jacquemart, D., Kleiner, I., Lacome, N., La_erty, W. J., Mandin, J.- Y., Massie, S. T., Mikhailenko, S. N., Miller, C. E., Moazzen-Ahmadi, N., Naumenko, O. V., Nikitin, A. V., Orphal, J., Perevalov, V. I., Perrin, A., Predoi- Cross, A., Rinsland, C. P., Rotger, M., _Sime_cov_a, M., Smith, M. A. H., Sung, K., Tashkun, S. A., Tennyson, J., Toth, R. A., Vandaele, A. C., and Vander Auwera, J.: The HITRAN 2008 molecular spectroscopic database, Journal of Quantitative Spectroscopy and Radiative Transfer, 110, 533{572, doi:http://dx.doi.org/10.1016/j. jqsrt.2009.02.013, URL http://www.sciencedirect.com/science/article/pii/ S0022407309000727, 2009.  R.A. Serway, J.W. Jewett.Principles of physics: a calculus-based text (4th edition), 809  Schmidt, G. A., Ruedy, R. A., Miller, R. L., and Lacis, A. A.: Attribution of the present-day total greenhouse e_ect, Journal of Geophysical Research: Atmospheres, 115, D20 106, doi:10.1029/2010JD014287, URL http://dx.doi.org/10. 1029/2010JD014287, 2010.  Seifritz, W.: Der Treibhause_ekt. Technische Ma_nahmen zur CO2- Entsorgung, Hanser, M•unchen, 1991.  Schneider, M. and Hase, F.: Ground-based FTIR water vapour profile analyses, Atmospheric Measurement Techniques, 2, 609–619, https://doi.org/10.5194/amt-2-609-2009, http://www.atmos-meas- tech.net/2/609/2009/, 2009.  Schmidt, M. W. I., et al., 2011: Persistence of soil organic matter as an ecosystem property. Nature, 478, 49–56.  Sepúlveda, E., Schneider, M., Hase, F., García, O. E., Gomez-Pelaez, A., Dohe, S., Blumenstock, T., and Guerra, J. C.: Long-term validation of tropospheric column-averaged CH4 mole fractions obtained by mid-infrared ground-based FTIR spectrometry, Atmospheric Measurement Techniques, 5, 1425–1441, https://doi.org/10.5194/amt-5-1425-2012, http://www.atmos-meas- tech.net/5/1425/2012/, 2012.  Tallis, L., Coleman, M., Gardiner, T., Ptashnik, I., and Shine, K.: Assessment of the consistency of H2O line intensities over the nearinfrared using sun- pointing ground-based Fourier transform spectroscopy, Journal of

172

Quantitative Spectroscopy and Radiative Transfer, 112, 2268–2280, https://doi.org/10.1016/j.jqsrt.2011.06.007, 2011.  Tropospheric Remote Sensing from Space John P. Burrows, Ulrich Platt and Peter Borrell - - Publisher: Springer Verlag, Heidelberg Springer Book Web Page Springer on line Page for the Book ISBN 978-3-642-14790-6 DOI 10.1007/978-3-642-14791-3  U.S. Global Change Research Program (2009) Global Climate Change Impacts in the United States.  U.S. Global Change Research Program (2000) Climate Change Impacts on the United States, The Potential Consequences of Climate Variability and Change.  UN Environment Programme (UNEP) and GRID-Arendal (2005) Vital Climate Change Graphics.  V. Saptari, Fourier-Transform Spectroscopy Instrumentation Engineering, SPIE Publication, Bellingham, 2003

 Wunch, D., Toon, G. C.,Wennberg, P. O.,Wofsy, S. C., Stephens, B. B., Fischer, M. L., Uchino, O., Abshire, J. B., Bernath, P., Biraud, S. C., Blavier, J.-F. L., Boone, C., Bowman, K. P., Browell, E. V., Campos, T., Connor, B. J., Daube, B. C., Deutscher, N. M., Diao, M., Elkins,  J. W., Gerbig, C., Gottlieb, E., Griffith, D. W. T., Hurst, D. F., Jiménez, R., Keppel-Aleks, G., Kort, E. A., Macatangay, R., Machida, T., Matsueda, H., Moore, F., Morino, I., Park, S., Robinson, J., Roehl, C. M., Sawa, Y., Sherlock, V., Sweeney, C., Tanaka, T., and Zondlo, M. A.: Calibration of the Total Carbon Column Observing Network using aircraft profile data, Atmospheric Measurement Techniques, 3, 1351–1362, https://doi.org/10.5194/amt-3- 1351-2010, http://www.atmos-meas-tech.net/3/1351/2010/, 2010.  Yang, J., Q. Liu, S.-P. Xie, Z. Liu, and L. Wu, 2007: Impact of the Indian Ocean SST basin mode on the Asian summer monsoon. Geophys. Res. Lett., 34, L02708.  Yang, X. C., Y. L. Hou, and B. D. Chen, 2011: Observed surface warming induced by urbanization in east China. J. Geophys. Res. Atmos., 116, 12.  Yang, Z., Wennberg, P., Cageao, R., Pongetti, T., Toon, G., and Sander, S.: Groundbased photon path measurements from solar absorption spectra of the O2 A-band, Journal of Quantitative Spectroscopy and Radiative Transfer, 90, 309 – 321, doi: http://dx.doi.org/10.1016/j.jqsrt.2004.03.020, 2005.  Yang, Z., Toon, G. C., Margolis, J. S. & Wennberg, P. O. 2002 Atmospheric CO2 retrieved from ground-based near IR solar spectra. Geophys. Res. Lett. 29, 531. (doi:10.1029/2001GL014537)  Wang, C., 2004: A modeling study on the climate impacts of black carbon aerosols. J. Geophys. Res., 109, D03106, doi:10.1029/2003JD004084.

173

 Washenfelder, R. A., Wennberg, P. O. & Toon, G. C. 2003 Tropospheric methane retrieved from ground-based near-IR solar absorption spectra. Geophys. Res. Lett. 30, 2226. (doi:10.1029/ 2003GL017969)  Wunch, D., C Toon, G., L Blavier, J.-F., A Washenfelder, R., Notholt, J., J Connor, B., W T Griffith, D., Sherlock, V., and Wennberg, P.: The Total Carbon Column Observing Network, 369, 2087–112, 2011. Wunch, D., Toon, G. C., Sherlock, V., Deutscher, N. M., Liu, X., Feist, D. G., and Wennberg, P. O.: The Total Carbon Column Observing Network’s GGG2014 Data Version, https://doi.org/10.14291/TCCON.GGG2014.DOCUMENTATION.R0/1221  Wunch, D., Wennberg, P. O., Toon, G. C., Keppel-Aleks, G. & Yavin, Y. G. 2009 Emissions of greenhouse gases from a North American megacity. Geophys. Res. Lett. 36, L15810. (doi:10.1029/2009GL039825)62, 2015.  Wunch, D., Toon, G. C., Wennberg, P. O., Wofsy, S. C., Stephens, B. B., Fischer, M. L., Uchino, O., Abshire, J. B., Bernath, P., Biraud, S. C., Blavier, J.-F. L., Boone, C., Bowman, K. P., Browell, E. V., Campos, T., Connor, B. J., Daube, B. C., Deutscher, N. M., Diao, M., Elkins, J. W., Gerbig, C., Gottlieb, E., Gri_th, D. W. T., Hurst, D. F., Jim_enez, R., Keppel-Aleks, G., Kort, E. A., Macatangay, R., Machida, T., Matsueda, H., Moore, F., Morino, I., Park, S., Robinson, J., Roehl, C. M., Sawa, Y., Sherlock, V., Sweeney, C., Tanaka, T., and Zondlo, M. A.: Calibration of the Total Carbon Column Observing Network using aircraft pro_le data, Atmospheric Measurement Techniques, 3, 1351{1362, https://doi.org/10.5194/amt-3- 1351-2010, 2010.  Wunch, D., C Toon, G., L Blavier, J.-F., A Washenfelder, R., Notholt, J., J Connor, B., W T Gri_th, D., Sherlock, V., and Wennberg, P.: The Total Carbon Column Observing Network, 369, 2087{112, https://doi.org/10.1098/rsta.2010.0240, 2011.  Wunch, D., Toon, G. C., Sherlock, V., Deutscher, N. M., Liu, X., Feist, D. G., and Wennberg, P. O.: The Total Carbon Column Observing Network's GGG2014 Data Version, https://doi.org/10.14291/TCCON.GGG2014.DOCUMENTATION.R0/1221662, 2015.  Source : XCO2 in an emission hot-spot region: the COCCON Paris campaign 2015 Felix R. Vogel, Matthias Frey , Johannes Staufer , Frank Hase , Grégoire Broquet, Irène Xueref-Remy , Frédéric Chevallier , Philippe Ciais , Mahesh Kumar Sha, Pascale Chelin , Pascal Jeseck , Christof Janssen , Yao Té , Jochen Groß , Thomas Blumenstock , Qiansi Tu, and Johannes Orphal

 Yang, Z., Washenfelder, R. A., Keppel-Aleks, G., Krakauer, N. Y., Randerson, J. T., Tans, P. P., Sweeney, C., and Wennberg, P. O.: New constraints on Northern Hemisphere growing season net ux, Geophysical Research Letters, 34, https://doi.org/10.1029/2007GL029742, 2007.

174

 W.D. Perkins, "Fourier Transform-Infrared Spectroscopy”. Part 1. Instrumentation. Topics in Chemical Instrumentation. Ed. Frank A. Settle, Jr. Journal of Chemical Education, 63:1, January 1986: A5-A10.  Hassen Aroui, Johannes Orphal and Fridolin Kwabia Tchana (2012). Fourier Transform Infrared Spectroscopy for the Measurement of Spectral Line Profiles, Fourier Transform - Materials Analysis, Dr Salih Salih (Ed.), ISBN: 978- 953-51-0594-7  Ref.1.1--http://www.odec.ca/projects/2008/chan8k2/greenhouse_conclusion.html  https://www.atmos-meas-tech.net/9/2223/2016/  https://doi.org/10.5194/amt-6-3527-2013  https://www.atmos-meas-tech.net/6/3527/2013/  www.atmos-chem-phys.net/12/8763/2012/  www.atmos-meas-tech.net/5/2969/2012/doi:10.5194/amt-5-2969-2012  www.nasa.gov  https://meteo.geo.auth.gr/el/paroxi-dedomenon  http://www.intechopen.com/books/fourier-transformmaterials- analysis/fourier-transform-infrared-spectroscopy-for-the-measurement-of- spectral-line-profiles  https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistr y_Textbook_Maps/Supplemental_Modules_(Physical_and_Theoretical_Chem istry)/Spectroscopy/Vibrational_Spectroscopy/Infrared_Spectroscopy/How_a n_FTIR_Spectrometer_Operates  https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistr y_Textbook_Maps/Supplemental_Modules_(Physical_and_Theoretical_Chem istry)/Spectroscopy/Vibrational_Spectroscopy/Infrared_Spectroscopy/Infrare d%3A_Theory  http://www.atmos-meas-tech.net/4/1061/2011/  https://doi.org/10.5194/amt-8-301-2015  http://www.atmos-meas-tech. net/8/301/2015/, 2015.  www.coolcosmos.ipac.caltech.edu: Hershel discovers infrared light  http://www.intechopen.com/books/fourier  www.wikipedia.org/Fourier transform infrared spectroscopy  http://www.crcpress.com  https://www.bruker.com/products/infrared-near-infrared-and-raman- spectroscopy/remote-sensing/em27sun/overview.html  https://www.selectscience.net/products/em-27-sun/?prodID=196076  https://www.epa.gov/climate-indicators/greenhouse-gases  https://slideplayer.com/slide/5905004/  www.explainthatstuff.com  https://www.ipcc.ch/reports/  https://hitran.org/

175

 https://www.ncdc.noaa.gov/monitoring-references/faq/greenhousegases.php  https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions  http://www.ghgonline.org/otherco.htm  http://www.climate-change-knowledge.org/ghg_sources.html  https://sentinel.esa.int/web/sentinel/user-guides/sentinel-5p-tropomi  http://www.tropomi.eu/  https://sentinels.copernicus.eu/web/sentinel/missions/sentinel-5p  (https://en.wikipedia.org/wiki/Photosynthesis)  https://www.sciencedirect.com/science/article/pii/B9780444637765000012

176

177