EATS 4230 “Remote Sensing of the Atmosphere”

The Need for Remote Sensing of the Atmosphere

In these days of ‘global change’ there is an ever increasing need for global measurements of key atmospheric ‘state parameters’. The most important being:

- Global Temperature fields

- Global Wind fields

-Minor constituent concentration fields, e.g. H2O, O3,

and anthropogenic {CFC’s, NO2, OClO, BrO, ClO} O3 depleters

These are important for monitoring global change, but they are also important for refining our understanding of atmospheric processes in general:

- Knowledge of T and wind fields enhances our understanding of atmospheric dynamics.

- Which in turn helps us understand the relative importance of transport and local chemical processes in influencing, say O3.

-Also required to isolate natural secular changes from anthropogenic effects. Which changes are natural and which are manmade?

For example:

We have some understanding of winds in the troposphere and the general circulation of the atmosphere below the tropopause – this largely comes from balloon releases at global stations. But we have very little observational data above about 10 km in altitude – and less well developed and tested theories. We know very little about winds in the upper atmosphere and how they vary with ‘time of day’, latitude, season & solar cycle.

These questions would best be answered if we could measure:

-Temperatures at all latitudes, longitudes, altitudes and local times

-Constituents at all latitudes, longitudes, altitudes and local times

-Winds at all latitudes, longitudes, altitudes and local times

Obviously this is an impossible task!

For example consider just temp measurements. We cannot just hang a thermometer at all locations, all heights constantly to create our global/temporal maps of temperatures. This is where ‘remote sensing’ comes in. Remote Sensing involves making measurements at a distance rather than directly at the point of interest – or in situ.

But how can we measure temperature at point ‘A’ without having a thermometer there? How can we measure the wind direction without placing a windsock there?

Or generally, how do we measure atmospheric parameter ‘A’ at point ‘B’ by observing some physical quantity ‘C’ from point ‘D’?

This is the challenge of ‘atmospheric remote sensing’.

This challenge is best met by using ‘optical techniques’ i.e., methods that exploit electrometric radiation (EMR). EMR generally travels in straight lines – without wires!!

We use properties of the EMR measured at point ‘D’ to infer info about the state of the atmosphere at point ‘B’. Since we know where the EMR came from.

As we know, EMR spans a huge range of frequencies and wavelengths.

Radio: λ~ 1000 – 1m

Microwave: 1m – 1mm

IR: 1mm – 1 μm

Visible: 1 μm (0.7- 0.3)

UV: 0.3 μm

In this course we will focus on the visible/near UV regions. Each of these regions has particular advantages and technical limitations or disadvantages.

Furthermore, with each region, different properties of how the radiation interacts with the atmosphere will be exploited. There are summarized as in Fig 1.1

Within each of these techniques, however, we will have to deal with an inherent difficulty, known as the inverse problem.

Since we will rarely be making a ‘direct measurement’, e.g. temperature by thermometer, we will be inferring temperature for some other indirect measurement. Such as radiation intensity, this may not uniquely and unambiguously define temperature.

Many of the techniques we will look at will involve solving the inverse problem of going ‘back’ from its tracks to determine the properties of the dragon.

Through much of the course we will be looking at this sort of problem using remote sensing of atmospheric temperatures from space as an example. So it is pertinent for us to start our discussion by looking at the standard temperature structure of the atmosphere.

Fig. 1.3

We note that the temperature drops throughout troposphere until the tropopause at ~12km, then increases throughout the stratosphere until the stratopause which peaks at about 50 km. This is where most ozone exists and the increasing temperatures in the stratosphere are due to heating caused by ozone absorbing solar UV radiation.

Above the stratopause the temperature falls again in the mesosphere until the mesopause is reached then increases up to very large values (~1500K) in the thermosphere.

Figure 1.3 shows both altitude and pressure in millibars (1 atm = 1013mb, 1013 hPa). Density and pressure drop exponentially with altitude – density scale height distance to 1/e is approximately is 7km.

As we will find ourselves bouncing around between both pressure scales and altitude scales (even some very obscure scales later!!) it is important to keep some idea of pressure and altitude in one’s mind.

A good rule of thumb is that pressure drops about 1/10 every 15 km – i.e. at 45-50km p=1/10³ = 1mb.

Note at 90km i.e. 90/15=6 the pressure equals 0.001mb or about 1 millionth of the surface pressure!!

The atmospheric temperature profile is not fixed but varies in both predictable and also unpredictable ways from place to place – hence the need for remote sensing of temperature profiles.A more global picture of the average atmospheric temperature structure is shown in Fig 1.4

. Fig. 1.4

This shows the latitudinal variation of temperature on an August day in both the southern & northern hemispheres.

These temperatures were measured with instruments on two US satellites (Nimbus 5&6) using some of the techniques we are going to learn about.

A slice through this contour plot will give a profile as shown in Fig 1.3

Notice that temperature drops in both hemispheres up to the tropopause but the height of tropopause moves up as we go towards the more southern latitudes on this August (northern summer) day.

Also notice that the temperatures at the tropopause get colder as we move towards the winter pole – reflects the effect of less solar heating?

At the stratopause the temperature gradient is not very strong .

More curious are the temperatures near the mesopause – here things are the wrong way around and the summer side is colder than the winter side!

In order to really understand anomalies like this we really would like to have reliable info on the temperatures, winds and pressures and some of the quantities that remote sensing scientists are challenged to provide are shown in Fig 1.5

Fig. 1.5

Now the bottom panel on Fig 1.5 raises the question of radiation budgets – balance between radiant energy coming in from the sun and radiant energy escaping out to space from the top of the Earth’s atmosphere.

These quantities – the in and out – are closely balanced in energy but they have very different spectral characteristics – in fact one could say that they have nothing in common in spectral distribution – which as we shall see is something of a blessing for remote sensing.

The in and out distributions shown in Fig. 1.6 may be approximately with black body radiance curves for temperatures of about 6000 K (outer regions of sun) and 250 K (temperature near the stratopause).

Fig. 1.6

A closer look at the spectrum of the incoming solar radiation shown in Fig 1.7 reveals that a black body at ~ 6000 K only approximates the solar in coming spectrum.

Fig 1.7

Fig 1.7 slows the spectral energy distribution of the solar radiation at the top and bottom of the atmosphere.

At the top of the atmosphere a solar temperature of 5900 K is really quite good particularly at the longer wavelengths. But notice that the spectrum at the bottom of the atmosphere shows large gaps were energy has been absorbed by the atmosphere.

All of these notches in the spectrum are due to spectroscopic absorption by minor atmospheric species mostly O3, H2O, and CO2.

The selective absorption by these gases (which are described as minor in terms of their concentrations but should not be considered as minor in terms of the roles they play in the atmosphere) is exploited by some of the remote sensing techniques we will learn about since the depth of the absorption is a measure of the concentration of the absorber.

We will be discussing shortly why and how these various molecules absorb the radiation at these particular wavelengths but I will simply point out at this stage that the absorption by O3 and O2 at these wavelengths is due to what are called ‘electronic transitions’ of the molecules. The absorption “features” by H2O and CO2 in this spectral region are due to what are known as ‘vibrational’ transitions of the molecules.

Having looked at the incoming radiation spectrum and identified some features that might be exploited for remote sensing, let’s look in some more detail at the outgoing radiation.

Fig. 1.8

Fig 1.8 shows the spectrum of the radiance emerging from the top of the atmosphere, i.e. the outgoing radiation from Earth and its atmosphere, as measured with an instrument looking downwards from the Nimbus 4 satellite.

Here again we see that a blackbody radiation curve is only a crude approximation to the outgoing spectrum and there again appear to be notches ‘cut out’ of blackbody radiation curve.

At first sight one might think that these notches, which occur where CO2, O2 and H2O have vibrational absorption bands, simple reflect the same absorption phenomena on the outgoing radiation. This is only partly true because, as we shall see, in the infrared region absorption of radiation from the ground is not the only thing that is going on.

Absorption plays a big role but re-emission of the absorbed or thermal emission radiation by these molecules actually fills in the gaps to a very significant extent.

The depths of the notches are not simply related to the concentration of the absorbers, as it is in the incoming visible case. However as we shall see, because of this absorption and re-emission or radiative transfer, the strength of the emission actually tells us a great deal about the temperatures and temperature profiles in the atmosphere – precisely one of the things we want to measure!

Fig. 1.9

Fig. 1.9 shows the same over the Sahara, the Mediterranean and Antarctic. Obviously the radiances over the Sahara are greater than those over the Mediterranean which are greater than those over Antarctica.

Looks like expected temp of surface. Indeed away from absorption bands the curves do follow the black body curve for the surface temperatures. However, the CO2 absorption band near 650 cm^-1 appears as emission in the Antarctica case!!

Why is this?? Well the temperatures in some regions of the atmosphere over Antarctica are greater than the ground and the atmosphere actual adds to the emission.

In this case the brightness at the CO2 band is determined by the temperature at some particular height in the atmosphere. In fact in the other two cases the brightness at the CO2 band is also determined by the atmospheric temp at some particular height in the atmosphere.

Two of the problems we will deal with are (1) determining what that characteristic height is and (2) devising means for recovering the temps at other heights so that we can recover the temperature profile.

However, before we get around to that we are going to look in more detail at the nature of the absorption and emission processes in the atmospheric molecules. This brings us to some basic spectroscopy of atmospheric molecules.