Introduction

In this chapter, we complete our survey of the principal types of remote sensing instrument by discussing those active systems that make direct use of the backscattered power. Optical (lidar) systems are used for sounding clouds, aerosols and other atmospheric constituents, for characterising surface albedo, and for measuring wind speeds. These are discussed briefly in section 9.2. However, the bulk of this chapter is concerned with microwave (radar) systems.

In section 9.3 the ground-work established in chapter 3 is extended to a derivation of the radar equation, which shows how the power detected by a radar system is related to the usual measure of backscattering ability, the differential backscattering cross-section σ0. The remainder of the chapter discusses the main types of system that employ this relationship. The first and simplest is the microwave scatterometer (section 9.4), which measures σ0, usually only for a single region of the surface but often for a range of incidence angles. As described here, this is not an imaging system, although the distinction between microwave scatterometers and imaging radars is not a precise one.

The last two sections discuss the true imaging radars. Section 9.5 describes the side-looking airborne radar (SLAR), or real-aperture radar, which achieves a usefully high spatial resolution in one dimension by time-resolution of a very short pulse. Resolution in the perpendicular direction is achieved by using an antenna with a narrow beamwidth, namely a large antenna.

Definition and origins of remote sensing

‘Remote sensing’ is, broadly but logically speaking, the collection of information about an object without making physical contact with it. This is a simple definition, but too vague to be really useful, so for the purposes of this book we restrict it by confining our attention to the Earth's surface and atmosphere, viewed from above using electromagnetic radiation. This narrower definition excludes such techniques as seismic, geomagnetic and sonar investigations, as well as (for example) medical and planetary imaging, all of which could otherwise reasonably be described as remote sensing, but it does include a broad and reasonably coherent set of techniques, nowadays often described by the alternative name of Earth Observation. These techniques, which now have a huge range of applications in the ‘civilian’ sphere as well as their obvious military uses, make use of information impressed in some way on electromagnetic radiation ranging from ultraviolet to radio frequencies.

The origins of remote sensing can plausibly be traced back to the fourth century bc and Aristotle's camera obscura (or, at least, the instrument described by Aristotle in his Problems, but perhaps known even earlier). Although significant developments in the theory of optics began to be made in the seventeenth century, and glass lenses were known much earlier than this, the first real advance towards our modern conception of remote sensing came in the first half of the nineteenth century with the invention of photography.

Introduction

Chapters 5 to 7 considered passive sensors, detecting naturally occurring radiation. In this chapter and the next we shall discuss active sensors, which emit radiation and analyse the signal that is returned by the Earth's surface or atmosphere. We have already identified three possible classifications of remote sensing systems, distinguishing between passive and active and between imaging and non-imaging, as well as classifying them according to the wavelength of radiation employed. We can also classify active systems according to the use that is made of the returned signal. If we are principally concerned with the time delay between transmission and reception of the signal we shall call the method a ranging technique, whereas if we are also (or mainly) interested in the strength of the returned signal we shall call it a scattering technique. The distinction between the two cannot be made entirely rigorous, but it provides a useful way of thinking about active remote sensing systems. It is clear that ranging systems are simpler both to visualise and, because of their less stringent technical demands, to construct, and we shall therefore consider them first. In chapter 9 we discuss the scattering techniques.

Laser profiling

Laser profiling (or laser altimetry) is the simplest application of the lidar (LIght Detection And Ranging) technique. Conceptually, it is extremely straightforward. A short pulse of ‘light’ (visible or near-infrared radiation) is emitted towards the Earth's surface by the instrument, and its ‘echo’ is detected some time later.

In chapter 3, we discussed principally the interaction of electromagnetic radiation with the surface and bulk of the material being sensed. However, the radiation also has to make at least one journey through at least part of the Earth's atmosphere, and two such journeys in the case of systems that detect reflected radiation, whether artificial or naturally occurring. Each time radiation passes through the atmosphere it is attenuated to some extent. In addition, as we have already seen in section 3.1.2 and figure 3.4, the atmosphere has a refractive index that differs from unity so that radiation travels through it at a speed different from the vacuum speed of 299 792 458 m s−1. These phenomena must be considered if the results of a remotely sensed measurement are to be corrected for the effects of atmospheric propagation, or if they are to be used to infer the properties of the atmosphere itself. We have already considered them in general terms in discussing the radiative transfer equation (section 3.4). In this chapter, we will relate them more directly to the constituents of the atmosphere.

Composition and structure of the gaseous atmosphere

At sea level, the principal constituents of the dry atmosphere are molecules of nitrogen (about 78% by volume), oxygen (21%) and the inert gas argon (1%). There is also a significant but variable (typically 0.1% to 3%) amount of water vapour, often specified by the relative humidity H.

Introduction

In chapter 5 we discussed photographic systems, and although these provide a familiar model for many of the concepts to be addressed in this and subsequent chapters, they nevertheless stand somewhat apart from the types of system to be discussed in chapters 6 to 9. In the case of photographic systems, the radiation is detected through a photochemical process, whereas in the systems we shall now consider it is converted into an electronic signal that can be detected, amplified and subsequently further processed electronically. This clearly has many advantages, not least of which is the comparative simplicity with which the data may be transmitted as a modulated radio signal, recorded digitally and processed in a computer.

In this chapter, we shall consider electro-optical systems, interpreted fairly broadly to include the visible, near-infrared and thermal infrared regions of the electromagnetic spectrum. The reason for this is a pragmatic one, since many instruments combine a response in the visible and near-infrared (VIR) region with a response in the thermal infrared (TIR) region, and much of the technology is common to both. Within this broad definition we shall distinguish imaging systems, designed to form a two-dimensional representation of the two-dimensional distribution of radiance across the target, and systems used for profiling the contents of the atmosphere. It is clear that an imaging system operating in the VIR region has much in common with aerial photography, and systems of this type are in very wide use from both airborne and spaceborne platforms.

Introduction

In this chapter we consider aircraft and satellites as platforms for remote sensing. There are other, less commonly used, means of holding a sensor aloft, for examples towers, balloons, model aircraft and kites, but we will not discuss these. The reason for this, apart from their comparative infrequency of use, is that most remote sensing systems make direct or indirect use of the relative motion of the sensor and the target, and this is more easily controllable or predictable in the case of aircraft and spacecraft. Figure 10.1 shows schematically the range of platforms, and their corresponding altitudes above the Earth's surface.

The spatial and temporal scales of the phenomenon to be studied will influence the observing strategy to be employed, and this in turn will affect the choice of operational parameters in the case of an airborne observation or of the orbital parameters in the case of a spaceborne observation.

Aircraft

Aircraft of various types provide exceptionally convenient and operationally flexible platforms for remote sensing, carrying payloads ranging from a few tens of kilograms to many tonnes. With a suitable choice of vehicle, a range of altitudes can be covered from a few tens of metres, where atmospheric propagation effects are negligible, to many thousands of metres, above most of the Earth's atmosphere. The choice of flying altitude will obviously have an impact on the scale, spatial coverage and spatial resolution of the data collected.

Introduction

The general direction of this book has been to follow approximately the flow of information, from the thermal or other mechanism for the generation of electromagnetic radiation, to its interaction with the surface to be sensed, thence to its interaction with the atmosphere, and finally to its detection by the sensor. It is clear that that the information has not yet reached its final destination. Firstly, it is still at the sensor and not with the data user. Secondly, the ‘raw’ data will in general require a significant amount of processing before they can be applied to the task for which they were acquired.

In this chapter, we shall discuss the more important aspects of the processes to which the raw data are subjected. For the most part, it will be assumed that the data have been obtained from an imaging sensor so that the spatial form of the data is significant. The principal processes are transmission and storage of the data, preprocessing, enhancement and classification. The last three processes are generally regarded as aspects of image processing, a major field of study in its own right, and we shall not be able to do much more here than outline its general features. There are many books on the subject to which the interested reader may be referred, for example Mather (1987), Richards (1993) and Schowengerdt (1997).

Transmission and storage of data

It is clear that the data must be brought from the sensor to the place where they are to be analysed.

Introduction

The Global Positioning System (GPS) is a satellite-based positioning system operated by the United States Department of Defense. It has been operational since 1993, and although it is primarily intended for military use, it is, subject to one or two provisos, also available for world-wide civilian use. It is now widely used in remote sensing – in field work, to determine the position of training areas for image classification (see section 11.3.4.2), and also by some remote sensing satellites, for precise determination of the satellite's own position. This appendix therefore presents a brief introduction to GPS and its capabilities. A much fuller discussion is given by, for example, Leick (1995).

The fundamental idea behind GPS is not a new one. It is a radio-positioning system in which timing signals are transmitted at known times from a number of radio beacons at known locations. By measuring the times at which these signals are received, the distances to the various beacons can be calculated, and hence the position of the receiver can be deduced. What sets GPS apart from its predecessors is that the beacons are carried on satellites, providing genuinely global coverage.

Space segment

The ‘space segment’ of the GPS system consists of 24 satellites in circular orbits around the Earth. The semi-major axis (radius) of these orbits is about 26 600 km and the inclination is 55°, giving them a nodal period (see equation (10.13)) of 43 082 s, or exactly half a sidereal day.

Introduction

Aerial photography, as we remarked in chapter 1, represents the earliest modern form of remote sensing system. Despite the fact that many newer remote sensing techniques have emerged since the first aerial photograph was taken in 1858, aerial photography still finds many important applications, and there are many books that discuss it in more detail than will be possible in this chapter. The interested reader is referred, for example, to chapters 2 to 5 of Avery and Berlin (1992). Aerial photography is familiar and well understood, and is a good point from which to begin our discussion of types of imaging system. In particular, it provides a convenient opportunity to introduce some of the imaging concepts that will be useful in discussing some less familiar systems in later chapters.

Photography responds to the visible- and near-infrared parts of the electromagnetic spectrum. It is, in the context of remote sensing, a passive technique, in that it detects existing radiation (reflected sun- and skylight), and an imaging technique, in that it forms a two-dimensional representation of the radiance of the target area. In this chapter, we shall consider the construction, function and performance of photographic film, especially its use in obtaining quantitative information about the geometry of objects. The chapter then discusses the effects of atmospheric propagation, and concludes by describing the characteristics of some real instruments and giving a brief account of the applications of the technique.

There are many books that explain the subject of remote sensing to those whose backgrounds are primarily in the environmental sciences. This is an entirely reasonable fact, since those people continue to be the main users of remotely sensed data. However, as the subject grows in importance, the need for a significant number of people to understand not only what remote sensing systems do, but how they work, will grow with it. This was already happening in 1990, when the first edition of Physical Principles of Remote Sensing appeared, and since then increasing numbers of physical scientists, engineers and mathematicians have moved into the field of environmental remote sensing. It is for such readers that this book, like its first edition, has been written. That is to say, the reader for whom I have imagined myself to be writing is educated to a reasonable standard (although not necessarily to first degree level) in physics, with a commensurate mathematical background. I have, however, found it impossible to be strictly consistent about this, because of the wide range of disciplines within and beyond physics from which the material has been drawn, and I trust that readers will be understanding when they find the treatment either too simple or over their heads.

This book attempts to follow a logical progression, more or less following the flow of information from the remotely sensed object to the user of the data. The first four chapters lay the general foundations.

Introduction

In chapters 5 and 6 we considered passive remote sensing systems in which the diffraction resolution limit λ/D, while important, was not usually a critical parameter of the operation. In this chapter, we consider our last major class of passive remote sensing system, the passive microwave radiometer. This is a device that measures thermally generated radiation in the microwave (usually 5–100 GHz) region. As we discussed in section 2.6, the long ‘tail’ to the Planck distribution at relatively low frequencies means that measurable amounts of radiation are emitted even in this range of frequencies.

Because microwave wavelengths are so much greater than those of visible or even of thermal infrared radiation, the resolution limit plays a much more important role, and we shall need to give careful attention to the factors that determine it. The treatment that follows in this chapter is similar to that of Robinson (1994), and is expanded upon by Ulaby et al. (1981, 1982, 1986). Much of the technology and nomenclature of passive microwave radiometry was originally developed in the field of radio astronomy, and further details can also be found in works on that subject.

Antenna theory

Angular response and spatial resolution

As we have remarked before, electromagnetic radiation is detected through its influence on electrons, which are excited to higher energy states by the incident photons. The energy of a microwave photon is typically only a few microelectron-volts, which is too small to excite an electron across an atomic or molecular band-gap.