INTRODUCTION

What is Remote Sensing, anyway?? Remote sensing means...

The most obvious example is photography, but Remote Sensing also includes:

-airborne and satellite-borne imaging of the Earth and other planets (all astronomy is remote sensing)

-using sonar to study the ocean floor

-using ultrasound or cat-scans for medical research

REMOTE SENSING IS ONLY A TOOL! It is just one tool among many that geologists, urban planners, planetary scientists, foresters, etc. use when doing scientific research. It is important to remember that if you are going to do science in some field then you need to know that field, not just remote sensing techniques.

REMOTE SENSING OFFERS ADVANTAGES IN THREE MAIN WAYS:

1) Remote Sensing offers a synoptic (big-area) view that you cannot get from the ground. Having an image of an entire mountain range makes looking for patterns of faults, slopes, etc. much easier to identify.

2) Remote Sensing allows you to view and study areas that are either very remote or very dangerous (or both).

3) Remote Sensing can utilize wavelengths of light that the human eye is not sensitive to. Examples include ultra-violet and infra-red portions of the spectrum - many times there is very useful information at these wavelengths.

BUT REMOTE SENSING DOES NOT REPLACE OTHER AVENUES OF RESEARCH, SUCH AS FIELD WORK, ANALYSIS OF SAMPLES IN THE LAB, OR GEOPHYSICAL TECHNIQUES. REMOTE SENSING IS JUST ANOTHER METHOD THAT SCIENTISTS CAN USE FOR ANSWERING QUESTIONS.

In addition to having to know the discipline you are planning to apply remote sensing to, you need to know how this tool works, namely what exactly are you measuring? How does a bright object in a visible image differ from a bright object in a thermal infra-red image or a bright object in a radar image? This is important because remote sensing utilizes the energy that is received at a sensor after interacting with the surface you are studying.

There are two major types of remote sensing, passive and active.

Passive remote sensing means that the sensing device just sits there and collects energy. This energy may be:

-reflected solar energy

-solar energy that hit the surface, was absorbed, and then was re-radiated by the surface

-energy generated by the surface itself (e.g., thermal energy of a lava flow or forest fire)

Active remote sensing means that the energy source is artificial. Examples include radar and sonar where the energy is generated and sent by the remote sensing device, bounces off the surface being studied, and is then collected back at the sensor.

As noted above, photography is probably the most common type of remote sensing. Photography, however, differs fundamentally from digital remote sensing and it is important to keep the two techniques separate in your mind because of these differences.

A photograph is considered to be an "analog" image. Photographic film contains small crystals of silver iodide, and these undergo a chemical reaction when a photon of light hits them. Then, the film is processed - the grains that were hit by light and those that were not react differently during processing and thereby reproduce the image. With photography you can make the following example statement: "the lake is darker than the island".

Digital images, however, are very different. Here, there is a grid pattern of detectors that are sensitive to incoming energy. You can think of this battery of detectors as looking like a checker board. Each detector makes up an element of the whole picture (it is a picture element or "pixel"). Each detector is able to measure the amount of energy that is hitting it and this measurement has a numerical value. So instead of just telling you whether or not light hit (such as with photographic film), a digital detector tells you how much energy has hit. This means you can make the following statement: "the lake is darker than the island because its surface temperature is cooler... or its salinity is different... or its surface is smoother" (depending on what type of sensor you are using).

Another important advantage to digital images is that because each pixel has a numerical value, computers can perform mathematical operations on these numbers. For example you can subtract the pixel values in one image from those in another - detecting changes that may have occurred between the times of collection of the two images. You can also enhance the contrast of a digital image quite easily, by changing the values of the pixels (there will be more on this later).

Because of these fundamental differences, we will be very careful to refer to digital images as "images" and photographic products as "photos."

SCALE

Before we start looking at some examples, we need to discuss the concept of SCALE. Scale refers to the relationship between the size of an object in a photo (or map or image) and its size in real life. For example, if an airport runway is 5 cm long in an air photo, and you know it is 1 km long in real life, the photo scale is:

5 cm = 1 km or

5 cm = 1000 m = 100,000 cm or

1 cm = 20,000 cm

You can illustrate this in 4 different ways:

scale = 1:20,000 (not the best way)

scale = 1/20,000 (not the best way)

1 cm = 200 m (not the best way)

put in scalebar (good)

The first two are showing you the relationship between the size of something on the photo and its size in real life. A house that is 1 mm wide in the image is actually 20,000 mm (=20m) wide in real life. The third one is telling you that 1 cm on the photo represents 200 m in real life. And the fourth one is showing you graphically how long 1 km or 100 m (or whatever) is in the photo or image.

Why is the last one the best? Because what if somebody takes the photo, puts it on a Xerox machine, and makes a copy at 150% enlargement? Now, that house that was once 1 mm wide is now 1.5 mm wide. Remember, however, that the scale tells you the relationship between the image and real life. Has the house magically become 20,000x1.5 mm = 30 m wide? Of course not. The scale on this enlarged map will therefore no longer be correct. What about the scale bar, however? All it tells you is how long a certain length is in the photo. If you enlarge the photo by 150% then everything in the photo is 150% bigger and you want the scale bar to represent that - it will.

PHOTOGRAPHY

Photography is very useful in all types of Earth Science, and comes in many forms. For example, a geologist may snap a photo of a particular feature while doing field work - that counts as remote sensing. Or, s/he may be flying over the area in an airplane or helicopter and happen to take a photograph. Finally, vertical air photos of the area may have been taken and available at a library or commercial outlet. Note that all these types of photography can be in stereo, a topic we will cover in a bit.

It is vertical air photos, particularly when used to produce stereoscopic views, that we will mainly concentrate on. There are two ways in which air photos can be used for gaining knowledge about the Earth's surface. In most cases a scientist won't use only one or the other, but instead, use a combination.

AIR PHOTO INTERPRETATION: This is a qualitative approach. It involves examining photo (or image) data and by recognizing features within the scene, deriving conclusions about the nature of the area as well as processes that are going on in that area.

Air photo studies fall into two main categories that often blend together. The first is AIR PHOTO INTERPRETATION. This is qualitative studying of the features in an air photo and requires good knowledge of the discipline involved, for example, identifying then recognizing the significance of a particular geological structure. For example, if you as a geologist recognize that a sharp bend in a stream course is indicative of an active fault cutting across the fault, then you can say something about the direction in which the fault is moving, how much it has moved, etc. Notice that this requires you to have the geological knowledge to not only recognize the feature but to also interpret its meaning correctly.

The second type of study is called PHOTOGRAMMETRY. Photogrammetry is quantitative, and involves measuring things in air photographs, including slopes, areas, spacings, azimuths, heights, etc. It is possible, for example, to photogrammetrically measure the heights of a whole bunch of cliffs in an area but to not have any idea what their geological significance is. Of course the best thing to do is combine qualitative recognizing with qualitative measurements, and that is what most air photo geologists do.

As their name implies, vertical air photographs are taken from airplanes, looking straight down. The highest quality photos require the use of special mapping cameras and the simultaneous recording of airplane altitude, attitude, azimuth, as well as the date and time the photos were collected.

RADIAL DISTORTION

An air photograph is a 2-dimensional representation of the actually 3-dimensional surface of the Earth. Because of this there are some distortions that occur. The most obvious one is called radial distortion and it occurs because the only place in an air photo where you are looking vertically downward is the very center of the photo. This point is called the nadir point.

Everywhere else in the photo you are looking outward at some angle, and the light rays that pass from some object through the lens onto the photographic film will travel inward at an angle.

Take a look at the paths of light rays from the bases and tops of the buildings in this diagram. Notice that unlike the building in real life, where the base is directly under the top (and you wouldn't actually be able to see the base if you were looking straight down), in the air photo representation the base and top do not line up and you can see both. Notice also that the farther away from the nadir point the building is, the greater the separation there is between base and top of the building. This is radial distortion.

The amount of radial distortion also depends on the height of an object. Thus the higher a building (or hill or mountain) and the farther it is from the nadir point, the more it will appear to "lean outwards". This means that on an air photo you cannot measure distances on the photo and convert them to true distances, even if you know the scale of the photo.

In this diagram note that although the two buildings are the same distance from nadir, the taller (green) one is stretched out (i.e. distorted) more than the shorter (red) one.

MEASURING HEIGHTS

Here is an example of photogrammetry. One of the things that radial distortion allows you to do is determine the heights of objects in an air photo, provided that you know some things about how the photo was collected. Specifically, you need to know how high the plane was when the photo was taken. There are a number of ways to know this, including looking to see if the altitude data are actually printed on the photo! (but remember, the altitude tells you the height above sea level and what you need to know here is the height above the ground). If you know this height, here's how you can determine heights.

This diagram shows how you can figure out the true height (h) of an object - here a building - using only a single air photo. You need to know the height above the ground that the photo was taken from (H). Then, in the photo itself you measure the distance from the nadir point of the photo to the top of the object (r), and the length of the object in the photo (d). The formula is simple: h=(d/r)H

STEREO AIR PHOTOGRAPHS

One of the most powerful remote sensing tools available is stereo air photography. It is also a method that has been around for quite a while.

A single photo, obviously, is a 2-dimensional object, and just a single photo may contain sufficient geological information for a study of the area. However, WAY MORE information can be gathered if you can see the area in 3-dimensions. In fact, the effective spatial resolution of the photo is increased. This is because instead of only relying on tonal variations (light vs. dark), you can also see height variations. The combination is more powerful than either alone.

How do you see in 3-d anyway? Seeing in 3-d involves the perception of what is called "depth", namely that an object is some distance away or that two objects are different distances away.

-One way to get a sense of depth is to know the actual size of the object you are looking at. If you see a tour bus and it is really tiny then you know it must be far away.

-Another way is if there is some sort of atmospheric haze. Close objects will be clear while far objects will get progressively less distinct the farther away they are.

-Finally, you can see in stereo, and that will be the topic of the next discussion.

Stereo viewing requires two views, for example from your two eyes (or using two photographs). To see the effect of using two eyes vs. only one (and how 2 eyes give you depth perception), try the following experiment:

First, have your eyes at the level of a table but closed. Have someone place a coin on the table. Then open only one eye and try putting your finger directly on the coin.

Next, try the same experiment but open both eyes. See how much easier it is to put your finger directly onto the coin when you have two views and are therefore seeing in stereo.

In real life (as long as you have two eyes), the two views from the two eyes form an angle:

This angle alpha is called the absolute parallax. As an object gets farther away you can see that alpha gets smaller and smaller. The combination of human eyes+brain can determine alpha as long as it is greater than about 1 minute. There are 60 minutes in a degree meaning that one minute equals 0.0167 degree. Anything farther away (i.e. alpha < 1 minute) is perceived by human eyes as being at infinity.