CHAPTER 4
REMOTE SENSING
IN NATURAL HAZARD ASSESSMENTS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
A. OVERVIEW OF IMPORTANT REMOTE SENSING ATTRIBUTESB. AERIAL REMOTE SENSINGC. SATELLITE REMOTE SENSINGD. APPLICATIONS OF REMOTE SENSING TECHNOLOGY TO NATURAL HAZARD ASSESSMENTS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
SUMMARY
This chapter provides planners with an overview of remote sensing technologies and their general application in natural hazard assessments. Characteristics of both aerial and satellite remote sensing techniques and the role remote sensing can play in detecting and mitigating several natural hazards are highlighted. |
One of the most important tools available to the regional planner is the remote sensing of the environment. Not only is it very useful in the planning process in general, but it is also valuable in detecting and mapping many types of natural hazards when, as is often the case, detailed descriptions of their effects do not exist. If susceptibility to natural hazards can be identified in the early stages of an integrated development planning study, measures can be introduced to reduce the social and economic impacts of potential disasters.
All natural hazards are amenable in some degree to study by remote sensing because nearly all geologic, hydrologic, and atmospheric phenomena that create hazardous situations are recurring events or processes that leave evidence of their previous occurrence. This evidence can be recorded, analyzed, and integrated into the planning process.
Most remote sensing studies concerned with natural hazards have been about an area's vulnerability to a disaster, the monitoring of events which could precipitate a disaster, and the magnitude, extent and duration of a disaster. This chapter tells planners what types of remote sensing information are suitable for identifying and assessing particular natural hazards and where to look for it.
Since the existing remote sensing information may be inadequate for a planning task or phase, this chapter also provides guidelines on selecting and acquiring the appropriate data. Only those sensor systems that are deemed capable of making a insignificant contribution to the development planning process are discussed, with their specific applications to the assessment of each of several natural hazards. It is assumed that planners and other readers are already familiar with basic remote sensing technology and vocabulary. If further details of techniques and/or applications are required, near state-of-the-art information is available in Sabins (1986), Lillesand and Kiefer (1987), and ASP (1983). An excellent overview of satellite imaging systems and disaster management can be found in Richards (1982).
While both aerial and satellite remote sensing techniques are presented, emphasis is placed on satellite-derived sensing because the data provide the synoptic view required by the broad scale of integrated development planning studies. Aerial remote sensing data are useful to natural hazard management for focusing on priority areas, verifying small-scale data interpretations, and providing information about features that are too small for detection by satellite imagery, but extensive aerial surveys commonly exceed the budget constraints of a planning study and may also provide more information than is necessary, particularly during the early stages of the study.
A. OVERVIEW OF IMPORTANT REMOTE SENSING ATTRIBUTES
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
1. Scale2. Resolution3. Image Contrast4. Time Frame5. Remote Sensing Images and Maps6. Output Formats
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
Effective utilization of remote sensing data depends on the ability of the user to be accurate and consistent when interpreting photographs, images, graphs, or statistics derived from remote sensing sources. While most planners have been introduced to photo and image interpretation in their formal training, the best use of the data usually requires analysis by people with experience in landform analysis, such as geologists, physical geographers, foresters, etc. A relatively small investment in the services of an experienced interpreter may avoid needless delays and inappropriate use of remote sensing data. Whether or not the planner does his own interpretation, he should have a working knowledge of remote sensing techniques and the capability to assess the validity of an interpretation, as well as the ability to use the derived information.
The factors that determine the utility of remote sensing data in natural hazard assessments are scale, resolution, and tonal or color contrast. Other factors include area of coverage, frequency, and data cost and availability.
DEFINITIONS
Instruments which record electromagnetic radiation emitted or reflected from the earth can be mounted on aircraft or satellites. The former are called aerial or airborne remote sensors and the latter, satellite or spaceborne remote sensors. These instruments record data using optical, electro-optical, optical mechanical, or electronic devices. In this chapter, visual displays analogous to photographs, made with such processes as radar and thermal infrared scanning and produced on a medium other than film, are referred to by the general terms "image" or "imagery" |
1. Scale
The scale to which a photograph or image can be enlarged, with or without optical or computer enhancement, determines in what phase of the development planning study this information should be used. Presentations at scales of 1:500,000 or smaller are useful during the Preliminary Mission and certainly in Phase I, Development Diagnosis, when more detail is not necessary. Imagery at a scale of 1:250,000 or larger is required during the project formulation and feasibility study activities of Phase II when detail is more important and where certain, but less obvious, aspects of natural hazards must be defined. Frequently it is possible to detect natural hazard phenomena on a small scale photograph or image, but it is impossible to annotate it without enlargement to larger scales. Thus, it is necessary to use imagery at scales commensurate with the levels of detail required for the particular stage of the study, as well as the size of the study area itself. In addition, the larger the areal extent of change associated with a natural event, the more useful satellite imagery becomes.
2. Resolution
Scale is meaningless in the absence of adequate spatial resolution, the capability of distinguishing closely spaced objects on an image or a photograph. Image resolution is determined by the size and number of picture elements or pixels used to form an image. The smaller the pixel size, the greater the resolution. In photography, resolution is limited primarily by the film grain size, but lenses and other technical considerations play important roles.
In both cases, imagery and photography, separability between adjacent features plays a very important part in the identification process. Enlargements of photography or imagery cannot improve resolution but only the working space for the interpretation.
Spectral resolution also needs to be taken into consideration when selecting the type of data since different sensors are designed to cover different spectral regions. Spectral resolution refers to the band range or band width offered by the sensor. Figure 4-1 shows the spectral regions most commonly used in remote sensing. Most natural disasters involve spectral changes. Floods lead to significant spectral changes whereas earthquakes lead to little spectral variation due to less spectral contrast in relation to non-affected areas.
3. Image Contrast
The contrast between features on an image or photograph is a function of the sensor's ability to record the tonal or spectral content of the scene. Different spectral bands of sensing systems may exhibit strong or weak contrasts depending on the regions covered on the electro-magnetic spectrum and the surface viewed. For example, a given band may show little contrast between vegetation types in a forest environment but may show strong contrasts between rock types in an arid area. Hazardous areas such as earthquake fault zones or areas susceptible to landslides may be too small for some sensors, e.g., Advanced Very High Resolution Radiometer (AVHRR) imagery, but may be readily visible on imagery produced by other sensor systems, e.g., Landsat Thematic Mapper (TM). The heavily vegetated and cloud-covered terrain of tropical Latin America and the Caribbean is among the most difficult to interpret geologically, but expert interpreters can detect many natural hazards through physiographic analysis of radar data which can penetrate clouds.
Figure 4-1 - Electromagnetic regions most commonly used in remote sensing
When an image does not provide the detail, resolution, or contrast that is needed, there are several options available. Since the identification of all desired features by interpretation from one sensor is not always possible, a second, completely different type of sensor, or even a combination of sensors, may be needed. Digital data can be enhanced and/or manipulated by using techniques such as contrast stretching, false color composites, principal component analysis, filtering, and supervised and unsupervised classifications.
4. Time Frame
The temporal occurrences of natural events will also affect the utility of remotely sensed data. Certain sensors can detect a phenomenon quite readily although their repeat coverage is every 16 days (Landsat). A flood could easily occur and recede within this time frame. On the other hand, desertification of an area can be a long process and the utility of remotely sensed data could be great for monitoring these changes. Events which are seasonal, predictable, or highly correlated with other events are more likely to benefit from imagery than events which occur randomly such as earthquakes or tsunamis (see Chapters 8-12).
5. Remote Sensing Images and Maps
To derive the most benefit from the use of available remote sensing data, the planners should use all supporting information on the study area (see Appendix A). Maps are particularly helpful in interpreting remote sensing data. Topographic maps are foremost among maps which help clarify many terrain recognition ambiguities found on remote sensing images. Geological maps bring attention to formations conducive to particular types of hazards. This knowledge can assist in localization and the systematic search for these hazards. Soils maps can serve a similar purpose, but to a lesser extent. Finally, vegetation and land-use maps can provide information on the moisture content, underlying geologic formations, and types of soils present.
In summary, remote sensing imagery should be regarded as data available to assist the planner in the assessment of natural resource and natural hazard information throughout the development of a planning study. The meaning and value of remote sensing data is enhanced through skilled interpretation used in conjunction with conventionally mapped information and ground-collected data.
6. Output Formats
Output formats consist of different ways in which remote sensing data can be presented. Photographic data are usually used in a film positive format or as a photographic print. Film data and photographic prints can be scanned and converted into digital data by being recorded on a computer compatible tape (CCT). The main advantage of digital data is the fact that they can be quantified and manipulated using various image processing techniques. Satellite or other images recorded on a CCT can be presented in a film positive format or photographed directly from the display monitor.
B. AERIAL REMOTE SENSING
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
1. Aerial Photography2. Radar3. Thermal Infrared Scanners4. Advantages and Limitations of Photography, Radar, and Thermal IR Scanners
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
Aerial remote sensing is the process of recording information, such as photographs and images, from sensors on aircraft. Available airborne systems include aerial cameras, multispectral scanners, thermal infrared (IR) scanners, passive microwave imaging radiometers, and side-looking airborne radars (SLAR). The systems offering the most practical and useful data in the context of integrated development planning and natural hazard assessments are aerial cameras, multispectral scanners, and thermal IR scanners and SLAR. This section describes the characteristics of the photography or imagery obtained from these three systems.
Availability of aerial remote sensing imagery varies for the type of data required. Aerial photography is readily available for many areas of study in most parts of the world, although in some instances it must be declassified for non-military use by the government of the country involved in the study. Radar imagery is also frequently classified.
Acquisition of infrared (IR) and radar data is more complex than aerial photography, although for a large area, radar may be less expensive than photography. Due to the specialized systems and operators required to produce IR and SLAR imagery, such data are usually available only from a limited number of organizations which either own or lease the systems. The cost to mobilize aircraft, equipment, and crews is high, but the cost of data coverage per line kilometer or per unit area can be reasonable if the area to be flown is large.
In addition to the type, availability, and cost of data, the planner should consider the conditions under which the acquisition of the appropriate data is taking place. Each sensor type has an optimum time of day, season, and/or table of appropriate conditions under which the best results are obtained. Also, to establish the current status of a hazard such as a volcano's activity, the interpretation of thermal IR imagery must be made close to the time of acquisition, and anomalies should be checked out immediately to determine the magnitude of temperatures that correlate with them. Currently obtained data, flown under similar instrument, weather, and terrain conditions, may be used to compare temporal variations of the hazard. In this way thermal pattern changes may be determined.
Thermal IR imagery information is the most transitory of any sensor data. There is a procession of changes in the thermal contrasts between the different materials on the ground, both terrain and vegetation. The transitions occur over daily and seasonal cycles and are modified considerably by the weather, soil, climate, relief, slope direction, and land-use practices. In spite of these masking variations, the thermal contrasts resulting from volcanic and geothermal activity can be interpreted by an experienced thermal IR interpreter.
The primary utility of SLAR imagery is in the interpretation of the relatively unchangeable elements of basic geologic structure and geomorphologic conditions. As a result, it is useful in studying many features related to natural hazards. Special SLAR image data acquisition is not normally feasible in a planning study budget, but previous coverage of the study area may be available. If it does exist, it should be sought and used to its fullest extent.
Both IR and SLAR imagery can be used in a stereoscopic mode but only where adjacent flight lines overlay. Since distortion due to air turbulence and/or differential altitude occurs during the raster-like development of each image as the aircraft moves forward, the stereoscopic model is imperfect. Despite these distortions the stereoscopic dimension is definitely an asset in helping to define natural hazards.
1. Aerial Photography
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
a. Scales and Wavelengthsb. Type of Film
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
Of all the sensors, aerial photography gives the closest representation to what the human eye sees in terms of wavelength response, resolution, perspective, stereoscopic viewing, and tonal or color values. The interpreter familiar with photographs can easily interpret these scenes, whereas other sensors, such as thermal IR scanners and SLAR systems, produce imagery whose appearance and physical basis is completely foreign to the inexperienced eye. Aerial photographs are probably the remote sensing data source with which the planner is most familiar (OAS, 1969).
a. Scales and Wavelengths
The most useful scales for aerial photographs range from 1:5,000 to 1:120,000. The need for reconnaissance type of information over large areas limits the use of photographs to the scales of 1:40,000 or less.
Photography is limited to the optical wavelengths which are composed of ultraviolet (UV), visible, and near-IR portions of the electromagnetic spectrum (see Figure 4-1). The first and last of these portions are recoverable on film under special film-filter conditions. The near-IR wavelengths are the reflective part of the larger infrared portion, which also includes emitted or thermal wavelengths.
b. Type of Film
Aerial photography may be obtained using black and white film, the least expensive medium, or with conventional color or color IR film. The type of film that should be used depends on its utility for a particular terrain being studied and the cost of the film. The speed of the film is also an important factor: the slower color films may not be used where the terrain is too dark, such as areas of ubiquitous heavy vegetation or predominantly dark rocks.
The two general types of black and white films used most frequently are the panchromatic and IR-sensitive films. Panchromatic films, which are negative materials having the same approximate range of light sensitivity as the human eye, are regarded as the standard film for aerial photography. It is the least expensive medium for aerial mapping and photo interpretation, but it may not be the logical choice for a given study area.
Black and white IR-sensitive film, although not commonly used, is a better choice for the penetration of strong haze and/or lush vegetation in humid tropical areas. It renders surface water, moisture, and vegetation contrasts much better than the standard film, and, as a result, can be an effective tool in regional planning and natural hazard assessments in humid tropical areas. There is, however, a diminution of detail in shadowed areas since scattered cooler light (blue end) is filtered out.
In high relief areas, it is best to shoot close to midday using IR films. In areas of low relief, photographs should be taken when the sun is low on the horizon (10° to 30°), causing shadows on the fine-textured surface. Low-sun-angle photography (LSAP) emphasizes textural characteristics of particular rock types, discontinuities, and the linear topographic features associated with faults and fractures. Vegetation types, both natural and cultivated, can also be defined to a large extent on a textural basis, and this may provide further information on the terrain. Almost any state-of-the-art aerial camera can capture LSAP using panchromatic or red-filtered infrared film.
The use of color films for natural hazard assessment takes various forms: negative film from which positive color prints are made, and positive transparencies, including color slides. To a limited extent, the negative films can be printed to emphasize certain colors and offer the ease of handling of prints. They do not have, however, the sharpness and dynamic color range of the positive transparencies, which are significantly better for interpretation purposes.
There are two major spectral types of color film: the natural or conventional color film, which covers the visible spectrum, and color IR film (green through near IR). The former is available as a negative (print) film and positive transparency, and the latter is available only as a positive transparency.
The IR color film response is superior to that of natural color films for a number of reasons. First, the yellow filter required for its proper use eliminates blue light that is preferentially scattered by the atmosphere. Eliminating much of the scattering greatly improves the contrast. Second, the differences in reflectance within vegetation types, soils, and rocks are commonly greater in the photographic IR component of this film. Third, the absorption of infrared and much of the red wavelengths by water enables a clearer definition of bodies of water and areas of moisture content. And fourth, the diminution of scattered light in shadowed areas enhances relief detail, thus improving the interpretation of the geomorphology. In view of these attributes, color IR film is preferred if color aerial photography is desired for humid tropical climates.
2. Radar
Radar differs from aerial photography as an aerial remote sensor. Unlike photography, which is a passive sensor system using the natural reflection from the sun, radar is an active sensor that produces its own illumination. Radar illuminates the terrain and then receives and arranges these reflective signals into an image that can be evaluated. These images appear similar to black and»d white photographs. The best use of airborne radar imagery in the development planning process and natural hazard assessments is the identification of geologic and geomorphologic characteristics. Radar imagery, like photography, presents variations in tone, texture, shape, and pattern that signify variations in surface features and structures. Of these elements, tonal variations which occur in conventional aerial photographs are the same as the eye sees. The tonal variations, which occur in radar images and appear as unfamiliar properties, are the result of the interaction of the radar signal with the terrain and vegetation. Just as it is not essential to fully understand the optical theory and processes involved with photography to be able to use aerial photographs, it is also possible to use radar images without a thorough understanding of electromagnetic radiation.
However, an interpreter needs to know something about how the image is formed in order to interpret it correctly and to appreciate fully the potential and limitations of radar. A skilled interpreter need only become familiar with the parameters that control radar return, understand their effect on the return signal, and recognize the effect of the side-looking configuration of the sensor on the geometry of the return signal.
Many useful radar images have been acquired in X-band, K-band, and Ka-band wavelengths (see Figure 4-2). However, X-band airborne radar systems are currently the most commonly offered by commercial contractors. In this band-width there are two basic types of systems: real aperture radar (RAR) and synthetic aperture radar (SAR). Real aperture or "brute force" radar uses an antenna of the maximum practical length to produce a narrow angular beam width in the azimuth (flight line) direction. The longer the antenna. the narrower the azimuth beam. A typical length is 4.5m, which approaches a maximum practical size for aircraft. For this reason the SAP was developed. The SAR is capable of achieving higher resolution without a physically large antenna through complex electronic processing of the radar signal.
The resulting resolution, coupled with the small scales at which images can be acquired, makes radar more suitable than photographic surveys for covering large areas. While RAR has a simple design and does not require sophisticated data recording and processing, its resolution in the range direction is relatively limited in comparison with the SAR of the same waveband. SAR maintains its high resolution in the range direction at long distances as well as its azimuth resolution. Resolution with SAR approaches 10m in azimuth and range.
3. Thermal Infrared Scanners
An airborne electro-optical scanner using a semiconductor detector sensitive to the thermal IR part of the spectrum is the best way to produce imagery that defines the thermal pattern of the terrain. Alternative methods using a television-like presentation have inadequate spatial resolution and thus cannot be used effectively from aircraft altitudes. They also lack adequate thermal resolution.
Figure 4-2 RADAR WAVELENGTH AND FREQUENCIES USED IN REMOTE SENSING FOR AIRCRAFT RADAR SYSTEMS
Band Designation a/
|
Wavelength (cm)
|
Frequency(v), GHz (109 cycles/sec-1)
|
Ka (0.86cm) |
0.8 to 1.1
|
40.0 to 26.5
|
K |
1.1 to 1.7
|
26.5 to 18.0
|
Ku |
1.7 to 2.4
|
18.0 to 12.5
|
X (3.0cm, 3.2cm) |
2.4 to 3.8
|
12.5 to 8.0
|
C |
3.8 to 7.5
|
8.0 to 4.0
|
S |
7.5 to 15.0
|
4.0 to 2.0
|
L (23.5cm, 25.0cm) |
15.0 to 30.0
|
2.0 to 1.0
|
P |
30.0 to 100.0
|
1.0 to 0.3
|
a/ Wavelengths commonly used in imaging radars in imaging radars are shown in parenthesesSource: Sabins, Floyd F., Jr. Remote Sensing: Principles and Interpretation (New York: W.H. Freeman, 1986).
Spatial resolution in scanners decreases with altitude above the terrain. Most commercial thermal infrared systems have spatial resolutions which provide for 2m to 2.5m resolution per 1,000m altitude at the nadir point (the point in the ground vertically below the camera) of the scan. Increasing the altitude above terrain to 2,000m would produce 4m to 5m spatial resolution.
Commonly, the 3.0 mm to 5.5 mm band provides the best information for "hot" objects (active volcanic vents, hot springs, etc.), while the 8.0 mm to 14.0 mm band provides the best information for features that are at ambient or cooler temperatures (flooding streams under canopies, warm springs, etc.). Frequently in studies involving IR surveys both bands are used to provide simultaneous imagery.
Properties of the airborne IR scanner system indicate that its practical use is restricted to the lower altitudes (under 3,000m) and, consequently, to relatively smaller areas than either radar or aerial photography. In natural hazard assessments, its best use would be in areas that are known or suspected to be areas of volcanism or where abnormal moisture conditions indicate dangerous situations. The latter may include, for example, trapping of water along active faults, or in back of landslide slumps, or moisture conditions associated with karst terrain.
IR scanning systems have drawbacks, but their unique capability of thermal imaging is unsurpassed. In addition, they can provide critical information for relatively small areas once specific hazard-prone areas have been identified.
4. Advantages and Limitations of Photography, Radar, and Thermal IR Scanners
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
a. Photography and Radarb. Thermal IR Scanners
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
a. Photography and Radar
Both aerial photography and radar have advantages and limitations. Photography cannot be used at any time in any weather as can radar. Radar can map thousands of square miles per hour at geometric accuracies conforming to national mapping standards. An area can be surveyed much more rapidly by radar than by aerial photography, and the final product provides an excellent synoptic view. Distance can be measured more accurately on radar than photography, and maps as large as 1:24,000 scale have been produced experimentally. The RADAM project of Brazil covered the country completely at a scale of 1:250,000. On the other hand, photography at the same scale shows considerably more detail, and it provides an excellent stereoscopic model for interpretation purposes in contrast to a more limited, but still useful, model obtained from radar. Aerial photography has the advantage of offering instantaneous scene exposures, superior resolution, ease of handling, and stereoscopic capability.
b. Thermal IR Scanners
Airborne electro-optical scanners, in general, can cover the electromagnetic spectrum using semiconductor electronic sensors from the UV through the visible and near IR into the thermal IR range of the spectrum. The utility of the UV spectrum in natural hazard and resource investigations has yet to be demonstrated, particularly when the image is degraded due to intense scattering of its rays. Scanners in the visible range are useful, especially when two or more wavebands are algebraically combined or manipulated.
Scanning imagery, because of its technique of recording a raster on film or tape, produces inherent distortions in the final built-up image scene. The lateral distortion from the flight line is reasonably corrected in the scanner system. Along the flight line, however, the rapid changes of altitude above the terrain during the formation of one scene from many scan lines produces many distortions. The persistent movement of the aircraft on three axes with limited stabilization presents the same problem. These distortions result in images that are difficult to interpret and whose location is difficult to identify, especially in mountainous and/or forested terrains. Despite these deficiencies, scanning from aircraft continues to be a very valuable method of obtaining thermal infrared imagery with reasonable spatial and thermal resolution.
In summary, aerial remote sensing provides information from aerial photographic cameras, side-looking radar, and thermal imaging scanners that is unsurpassed in resolution in their respective coverage within the electromagnetic spectrum. These systems produce imagery that ranges from the familiar visible spectrum to the unfamiliar infrared and microwave radar (short radio) spectra. This information can be used in conjunction with conventional maps of all kinds to enhance the data available to the planner.
C. SATELLITE REMOTE SENSING
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
1. Landsat2. SPOT3. Satellite Radar Systems4. AVHRR5. Metric Camera6. Large Format Camera7. Sojuzkarta
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
This section describes several satellite remote sensing systems which can be used to integrate natural hazard assessments into development planning studies. These systems are: Landsat, SPOT satellite (Système Probatoire pour l'Observation de la Terre), satellite radar systems, Advanced Very High Resolution Radiometer (AVHRR) on NOAA-10 and 11 satellites, metric camera, large format camera (LFC), and Sojuzkarta. Remote sensing from satellite vehicles has become increasingly important following the successful launch of the Landsat 1 satellite (formerly ERTS-1) in 1972. Since then many satellites with remote sensing capabilities have been developed and used successfully.
The Landsat multispectral scanner (MSS) provided the first practical imagery in four different spectral bands from space. The characteristics of this and other Landsat sensors are summarized in Figure 4-3. The accompanying return beam vidicon (RBV) sensor on this and subsequent satellites of this series were never noticed by scientists and planners like the MSS. The broad areal coverage of the Landsat sensors and the others that have followed, together with the capability to process the sensor data digitally, has made the satellite-derived data useful to regional planners and others interested in natural hazard assessments. The synoptic view of lands targeted for development can be imaged in an instant of time, Satellite imagery can provide a continuity of viewing conditions for large areas that is not possible on aerial photographic mosaics.
In addition to MSS, other satellite-borne sensors warrant discussion since they are potential tools for assessing natural hazards. Each sensor has its advantages and limitations in coverage of areas of interest and its resolution capability to define certain types of hazards. Some sensors are experimental, and provide limited areal coverage and lack temporal continuity. However, where coverage is available for a study area, the sensor data should be used in conjunction with existing data derived from Landsat or SPOT. The data derived can produce an inexpensive synergistic effect by combining data from more than one part of the spectrum, and are well worth the relatively small expense.
Ideally, it would be desirable to use a "multi-stage" approach to the natural resource and natural hazard assessments. This would involve using aerial photography and ground checks to establish more detailed knowledge of sample or representational sites. This may then be extrapolated over a larger area using Landsat or other satellite-derived data. Figure 4-3 shows the needed imagery characteristics for the assessment of various natural hazards-earthquakes, volcanic eruptions, landslides, tsunamis, desertification, floods, and hurricanes-for purposes of planning and mitigation. The characteristics of applicable satellite sensing technologies are described below.
1. Landsat
Since the Landsat series of satellites have been operational for a long period of time, there is a very large data base available, both in areal coverage and in repetitive coverage, through different seasons and during periods of natural disasters. Landsat MSS coverage exists from 1972 to the present in four spectral bands at 80m resolution. The thematic mapper (TM) was introduced on Landsat 4 in 1982, with seven spectral bands, six of them with 30m resolution and one in the thermal IR range with 120m resolution (see Figure 4-3).
Sensor data are digitally transmitted to ground stations in various parts of the world where they are recorded on magnetic tapes and preprocessed to improve their radiometric, atmospheric, and geometric fidelity. Ground receiving stations that cover Latin America and the Caribbean are in California, Maryland, Brazil, and Argentina. Distribution centers for Landsat sensor imagery are listed in the box below.
Although none of the existing satellites and their sensors has been designed solely for the purpose of observing natural hazards, the variety of spectral bands in visible and near IR range of the Landsat MSS, and TM and the SPOT HRV sensors provide adequate spectral coverage and allow computer enhancement of the data for this purpose. Repetitive or multitemporal coverage is justified on the basis of the need to study various dynamic phenomena whose changes can be identified over time. These include natural hazard events, changing land use-patterns, and hydrologic and geologic aspects of a study area.
The use of Landsat MSS and TM imagery in natural resource evaluations and natural hazard assessments is facilitated by the temporal aspect of available imagery. Temporal composites from two or more different image dates allow the recognition of hazard-related features that have changed, such as alteration of floodplains or stream channels and large debris slides, and to some extent, early recognition of disasters that evolve over time, such as drought or desertification. Chapter 8 has a detailed discussion of the use of Landsat sensors in flood hazard assessments. Specific manipulation and combination of the MSS or TM tape data of various bands of the same scene can increase the utility of the data.
Three-dimensional analysis, or stereoscopy, is essentially missing on the MSS and TM data. With MSS on Landsat 1, 2, and 3, there is an 18-day cycle, and sidelap is 14 percent at the equator, increasing poleward to 34 percent at latitude 40 and to 70 percent at polar latitudes. (Sidelap is the amount of overlapping of adjacent image coverage.) On Landsat 4 and 5, a lower altitude and a 16-day cycle with wider spacing results in only 7.6 percent sidelap at the equator and negligible increase toward the poles for both MSS and TM data. Unfortunately, the areas at the lower latitudes where our interests lie have the minimal stereoscopic coverage.
Figure 4-3 CHARACTERISTICS OF LANDSAT SENSORS
RBV |
1,2,3
|
PAN
| 0.505-0.750 |
920
|
79x59 d/
|
185x185 d/
|
every 18 days
|
1 b/
| 0.475-0.575 |
30x30 e/
|
99x99 e/
| ||||
2 c/
| 0.580-0.680 | ||||||
3 c/
| 0.690-0.830 | ||||||
MSS |
1,2,3,4,5
|
4 (green)
|
0.5-0.6 f/
|
920 g/
|
79x57 g/
|
185x185 f/
|
every 18 days g/
|
5 (red)
|
0.6-0.7
|
705h/
|
60x60h/
|
185x170 g/
|
every 16 days h/
| ||
6 (near IR)
|
0.7-0.8
|
237x237 i/
| |||||
7 (near IR)
|
0.8-1.1
| ||||||
8 (thermal)
|
10.4-12.6 b/
| ||||||
TM |
4,5
|
1
|
0.45-0.52
|
705
|
28.5x28.5 k/
|
85x170
|
every 16 days
|
2
|
0.52-0.60
|
120x120 l/
| |||||
3
|
0.63-0.69
| ||||||
4
|
0.76-0.90
| ||||||
5
|
1.55-1.75
| ||||||
6
|
10.40-12.50 i/
| ||||||
7
|
2.08-2.35
|
a/ RVB, Return Beam Vidicon; MSS, Multispectral Scanner; TM, Thematic Mapper; IR, Infrared.b/ Panchromatic and Landsat 3 onlyc/ Bands 1,2,3 on Landsat 1 and 2 onlyd/ Landsat 1 and 2e/ Landsat 3f/ Also called bands 1 to 4 on Landsat 4 and 5g/ Landsat 1 to 3h/ Landsat 4 and 5i/ Band 8 on Landsat 3j/ Thermalk/ Bands 1 to 5 and 7l/ Band 6 onlySource: Adapted from Budge, T. A Directory of Major Sensors and Their Parameters (Albuquerque, New Mexico: Technology Application Center, University of New Mexico, 1988).
If the terrain is flat with little relief, the little stereoscopic sidelap present would not be effective. In areas of rugged terrain, any small stereoscopic coverage would be welcome, especially if it fell within a critical part of a project area.
The return beam vidicon (RBV) is a framing camera system that operates as an instant television camera. It has not enjoyed the same popularity as the MSS, even though it provides useful information. Landsat 1 and 2 carried three RBVs that recorded green, red, and solar IR images of the same scene as the MSS did. They were capable of producing color IR images of 80m resolution, the same as the MSS, but were decidedly inferior due to technical problems. Landsat 3 carried an RBV system that acquired single black and white images in quadrants of the MSS scene in the 0.50mm to 0.75mm waveband, a spectral response from the green through red. The ground resolution was 40m, much better than the existing MSS and the earlier RBV resolution, enabling the recognition of natural hazard evidence of smaller scale.
The broad response of the RBV, however, did not enhance any particular feature or differentiate vegetation or rocks as well as the MSS bands. Its advantage lay primarily in providing higher spatial resolution for larger scale mapping of spectrally detectable features. In this regard it complemented the lower resolution MSS data which covered the same area. The RBV system was dispensed with entirely on Landsat 4 and 5, leaving only the MSS and TM sensors. The former was included to continue the temporal library with that type of sensor data and their 80m spatial resolution. The TM, with a 30m resolution, negated any requirement for the ineffective and little-used RBV system. Despite its absence on Landsat 4 and 5, RBV data of certain heavily vegetated tropical areas may be the only available source of data with adequate resolution for temporal comparison with later TM data.
SOURCES FOR LANDSAT IMAGERY Argentina Established: November 1980 Reception: MSS Distribution Center: Comisión Nacional de Investigaciones Espaciales Centro de Procesamiento Avenida Dorrego 4010 1425 Buenos Aires, Argentina Telephone: 722-5108; Telex: 17511 LANBAAR Brazil Established: May 1974 Reception and Processing: MSS and TM Distribution Center: INPE-DGI Caixa Postal 01, Cachoeira Paulista, SP CEP 12630, Sao Paulo, Brazil Telephone: (125) 611507; PBX: (125) 611377 Telex: 1233562 INPE BR United StatesEstablished: July 1972 Reception and Processing: MSS and TM Distribution Center: Earth Observation Satellite Company (EOSAT) 4300 Forbes Blvd., Lanham, Maryland 20706 Telephone: (301) 552-0500 or 800-344-9933 Telex: 277685 LSAAT UR |
The thermal IR portion of the TM was originally placed in the 10.4 mm to 12.5 mm spectral window where the earth's radiant energy is so low that a large detector is required. This resulted in a 120m ground resolution cell which generalized thermal detail, limiting its value for detecting the subtle and finely detailed geothermal changes associated with volcanic activity. The thermal resolution is 0.5° K (degrees Kelvin), which by airborne IR scanner standards (0.1° K or smaller),is poor. The best possible application in natural hazard assessments may be in active floodplain delineation and perhaps as a crude index to regional volcanic activity. The thermal infrared band (band 8) in Landsat 3 (10.4 mm - 12.5; mm with 240m spatial resolution) never worked properly and is, therefore, of no consequence for the applications discussed here. The blue-green band (0.45 mm - 0.52 mm) in the TM system (band 1) is unique among sensors aboard natural/resource-oriented satellites. The reason that this band has not been a part of the spectrum sought from satellites is the severe scattering of the blue light, which can badly degrade the image contrast where there is high humidity and/or high aerosol content in the atmosphere. However, in water, blue light has the best penetration capability in the visible spectrum.
In clear, sediment-free waters it can define sea bottoms to depths of 30 or more meters, depending primarily on the angle of incidence of the sun's illumination and the reflectance of the bottom. This property is useful for determining offshore slope conditions relevant to potential tsunami run-up.
Satellite remote sensing: approximate costs of basic data acquisition (CCTs, June 1989)
System |
Cost per km2
|
Landsat MSS |
US$.02
|
Landsat TM |
US$.11
|
SPOT |
US$.47-.61
|
Figure 4-4 CHARACTERISTICS OF SPOT SENSORS
SPOT SENSOR: MULTISPECTRAL HIGH-RESOLUTION VISIBLE (HRV)
| |||
Band
|
Wavelength (mm)
|
Resolution (m)
|
Image Format
|
XS1 |
0.50-0.59
|
20
| 60km swath with vertical |
XS2 |
0.61-0.68
|
20
| viewing angle and up to |
XS3 |
0.79-0.89
|
20
| 80km with +/- 27° viewing angle from vertical |
SPOT SENSOR: PANCHROMATIC HIGH-RESOLUTION VISIBLE (HRV-P)
| |||
Band
|
Wavelength (mm)
|
Resolution (m)
|
Image Format
|
P |
0.51-0.73
|
10
| 60km swath with vertical viewing angle and up to 80km with +/- 27° viewing angle from vertical |
2. SPOT
The SPOT satellite with its High Resolution Visible (HRV) sensors is similar in many respects to the Landsat satellite with its MSS and TM sensors. The HRV multispectral sensor (XS) ranges from the green wavelength into the near IR. The HRV-XS coverage is in three spectral bands rather than the four found in the MSS, but with much higher spatial resolution (20m versus 80m), although it covers only about 1/9 of the area covered by a Landsat scene. Additionally, SPOT carries a panchromatic sensor (HRV-P) which covers the green through red portions of the visible spectrum in a single band with 10m resolution. Both HRV sensors cover a 60km swath along the orbit path. It is possible to obtain simultaneous side-by-side coverage from each sensor, producing a 117km swath width, although this capability has not been frequently used. Figure 4-4 above summarizes the characteristics of SPOT sensors and their image formats.
The SPOT sensors have the unique capability of being pointable, 27 degrees to the left or right of the orbital track. This feature allows for repeated off-nadir viewing of the same ground swath, producing image stereopairs. The base-height ratios range from 0.75 at the equator to 0.50 at the mid-latitudes. This provides a strong vertical exaggeration. This third dimension, if it is available for a particular study area, together with the higher image resolution, can make SPOT'S sensors superior to those of Landsat if greater spectral resolution is not required. The sources for SPOT data are listed in the box above.
SOURCES OF SPOT IMAGERY
ARGENTINA:Comisión Nacional de Investigaciones Espaciales (CNIE) Av. del Ubertador 1513 Vicente López 1638 Buenos Aires, Argentina BOLIVIA: CIASER Casilla de Correo 2729 La Paz, Bolivia BRAZIL SENSORA Rua Bertolomeu Portela, 25 S/Lojas Botafogo, Rio de Janeiro CEP 2290, Brazil CHILE: S.A.F. Casilla 67 Correo Los Cerillos Santiago de Chile MEXICO: INEGI San Antonio Abad 124 Mexico City, 8 D.F., PERU: ONERN 355 calle 17, Urb. El Palomar, San Isidro Lima, Peru UNITED STATES: SPOT Image Corporation 1897 Preston White Drive Reston, Virginia 22091-4326, U.S.A. telephone: (703) 620-2200 VENEZUELA:CPDI Edo. Miranda Apartado 40200 Caracas, Venezuela 1040 A |
3. Satellite Radar Systems
There is considerable radar coverage throughout the world, and more space-derived radar data can be expected in the future.
The family of space radars stems from the Seasat (U.S.A.) radar, which was a synthetic aperture system that was especially designed for studying the ocean surface. In this capacity it had a large (70° average) depression angle to study the relatively flat ocean surface. For this reason Seasat's usefulness for imaging land extended to those land areas of low relief. During its short life in 1978, it managed to acquire a large amount of data from Western Europe, North and Central America, and the Caribbean.
Seasat was followed by Space Shuttle imaging radars SIR-A and SIR-B. Data from these radars was obtained from Space Shuttle flights in 1981 and 1984. Their characteristics along with those of Seasat are shown in Figure 4-5. SIR-A and SIR-B provided greater worldwide coverage, including large parts of Latin America, because the image data were recorded on board the Space Shuttle rather than telemetered to a limited number of receiving stations within range of the spacecraft, as was the case with the unmanned Seasat radar satellite.
Figure 4-5 - CHARACTERISTICS OF SEASAT, SIR-A, AND SIR-B SYSTEMS
Characteristics
|
Seasat (1978)
|
SIR-A (1981)
|
SIR-B (1984)
|
Repeat Coverage |
irregular, northern hemisphere
|
little to none
|
little to none
|
Resolution |
25x25m
|
40x40m
|
25x(17-58)m
|
Wavelength (23.5cm) |
L-band
|
L-band
|
L-band
|
Latitude coverage |
72°N-72°S
|
50°N-35°S
|
58°N-58°S
|
Altitude |
790km
|
250km
|
225km
|
Image-swath width |
100km
|
50km
|
40km
|
Source: Adapted from Budge, T. A Directory of Major Sensors and Their Parameters (Albuquerque, New Mexico: Technology Application Center, 1988).
SOURCES OF SATELLITE RADAR IMAGERY
SIR-A and SIR-B:National Space Science Center World Data Center A for Rockets and Satellites Code 601 NASA/Goddard Space Flight Center Greenbelt, Maryland 20771, U.S.A. Telephone: (301)286-6695 Seasat:NOAA, National Environmental Satellite Data and Information Service World Weather Building, Room 100 Washington, D.C. 20233, U.S.A. Telephone: (301)763-8111 |
The long wavelengths of these radar systems permit potential subsurface penetration between 2m and 3m in extremely dry sand (Schaber et al., 1986). There may be areas of hyperacidity in South America that may permit this type of penetration. This property may have some application to natural hazard assessment that is not readily apparent, as well as to integrated development planning studies. The problem seems to be that while significant amount of radar coverage is available, much has yet to be acquired where needed.
The SIR series of radar data acquisition is expected to continue with SIR-C in the future. Other radar sensors will be placed into orbit soon: Canada's Radarsat, a C-band (6.0cm) radar designed to provide worldwide stereoscopic coverage, is planned for the 1990s; the European Space Agency expects to launch a C-band synthetic aperture radar aboard the Earth Resources Satellite (ERS) in 1990; and Japan will launch an L-band imaging radar satellite in 1991. Thus, it can be expected that more radar imagery is forthcoming which will provide additional tools for natural hazard assessment.
4. AVHRR
The Advanced Very High Resolution Radiometer (AVHRR) on board the NOAA-7 through 11 satellites would not normally be considered useful for natural hazard assessments on the basis of its low resolution (1.1 km at nadir) alone. However, its large swath width of 2253km provides daily (day and night) coverage of the inhabited parts of the earth (see Figure 4-6). The near-nadir viewing repeat cycle is nine days, but the same area is still viewable from different angles within the swath from space. This results in complicated radiometric and geometric comparisons between different dates of data acquisition.
This scanning radiometer has 5 bands, which include band 1 (green to red), band 2 (red to reflected IR), band 3 (middle IR), band 4 (thermal IR), and band 5. The most useful bands are the thermal IR bands 4 and 5 particularly where moisture or ice is involved.
These have been successfully used to delineate flooded areas using temporal analysis techniques within 48 hours following a major flood event (Wiesnet and Deutsch, 1986). The thermal resolution in these bands is better than the Landsat TM thermal band 6 but with a significant trade-off loss in spatial resolution (1.1km versus 120m, respectively).
5. Metric Camera
The metric camera was an experiment on the STS-9/Spacelab 1 Mission in 1983 to determine whether topographic and thematic maps at medium scales (1:50,000 to 1:250,000) could be compiled from mapping camera images taken from orbital altitudes. Due to a late November launch date, illumination conditions were poor over many of the candidate target areas. As a result, slower shutter speeds had to be used than were planned, causing some image smear. Nevertheless, high quality images with a photographic ground resolution of about 20m were obtained on a 23cm x 23cm format panchromatic and color IR film. Analysis has shown that these images may be used for mapping at a scale of 1:100,000. On this mission, despite the fact that many problems were encountered, an area of more than 11 million km2 was covered. Plans are now underway to modify the camera to compensate for forward image motion and to refly it. It is expected that a ground resolution of about 10m would be obtained, permitting maps with a scale as large as 1:50,000 (Schroeder, 1986).
Figure 4-6. AVHRR CHARACTERISTICS
Platform: NOAA satellites (formerly Tiros)
| ||||
Spectral bands
|
Tiros-N
|
NOAA-6.8,10
|
NOAA-7,9,d,H,I,J
| |
1 |
0.55 - 0.90
|
0.58 - 0.68
|
0.58 - 0.68
| |
2 |
0.725 - 1.00
|
0.725 - 1.00
|
0.725 - 1.00
| |
3 |
3.55 - 3.93
|
3.55 - 3.93
|
3.55 - 3.93
| |
4 |
10.50 - 11.50
|
10.50 - 11.50
|
10.30 - 11.30
| |
5 |
none
|
none
|
11.50 - 12.50
| |
Altitude: 833-870km | ||||
Resolution: | Large Area Coverage (LAC): 1km | |||
Global Area Coverage (GAC): 4km | ||||
Image size: 2253 km swath | ||||
Repeat coverage: Daily, worldwide |
SOURCE OF AVHRR IMAGERY
Satellite Data Services DivisionNOAA/NESDIS/NCDC World Weather Building, Room 100 Washington, D.C. 20233, U.S.A. Telephone: (301) 763-8111 |
SOURCE OF METRIC CAMERA PHOTOGRAPHY
Deutsche Forschungs- und Versuchsanstalt furLuft- und Raumfahrt e.V. (DFVLR) Oberpfaffenhofen D-8031 Wessling Federal Republic of Germany |
Ground coverage of 190km x 190km per photograph frame was obtained using a 305mm lens from an altitude of 250km, yielding an image scale of 1:820,000. Overlap of 60 to 80 percent, obtained for topographic mapping purposes, is of great value in interpretation for natural hazards. The high resolution and stereocoverage make this photographic sensor system a potentially useful tool when sufficiently enlarged.
Five lines of metric camera photography cover parts of Latin America, and with resumption of the U.S. Space Shuttle program, additional high quality space photographs of areas of interest may become available.
6. Large Format Camera
The large format camera (LFC) photography was obtained during a Space Shuttle flight in October 1984. The term "large format" refers to the 23cm by 46cm film size, which was oriented with the longer dimension in the line of flight. LFC acquired 1,520 black and white, 320 normal color, and 320 IR color photographs, covering many areas within Latin America and the Caribbean. The scale of the photographs ranges from 1:213,000 to 1:783,000 depending on the altitude of the Space Shuttle, which varied between 239km and 370km. The swath covered a range between 179km and 277km, and each frame covered between 360km and 558km in the flight direction. Forward overlap up to 80 percent was obtained, allowing vertical exaggerations of 2.0, 4.0, 6.0, and 7.8 times in the stereomodels. Most photographs were taken with 60 percent overlap, which provided 4 times vertical exaggeration and an excellent stereomodel. The spatial resolution was about 3m for the black and white film and about 10m for the color IR film.
The availability of this excellent stereophotography, which can be enlarged ten times or more with little loss of image quality, is limited to certain areas covered by the ground track of the Space Shuttle. Some of this coverage includes clouds or heavy haze, but despite the limitations of coverage and occasional poor quality, the existing photography should be examined for its possible use in any regional natural hazards assessment and planning study.
Given the range of tools available for aerial and satellite remote sensing, their applications vary based on the advantages and limitations of each. The planner can regard each of these as a potential source of information to enhance natural resource evaluation and natural hazard assessment. The next section covers some of the applications of photographs and images in natural hazard assessments.
7. Sojuzkarta
Sojuzkarta satellite data consist of photographs taken with the KFA-1000 and KM-4 camera. Computer compatible tapes (CCTs) for digital image processing are not available, although it is possible to convert the data into digital format by using a scanner. Photographs obtained through the KFK-1000 camera have 5m resolution in the panchromatic mode and 10m resolution in the color mode; scales range from 1:220,000 to 1:280,000. KM-4 photography has a 6m resolution and is available at scales of 1:650,000 and 1:1,500,000. Applications of this sensor to natural hazard studies are likely to be in desertification monitoring, flood hazard and floodplain, and landslides studies.
SOURCES OF LFC PHOTOGRAPHY Chicago Aerial Survey, Inc. LFC Department 2140 Wolf Road Des Plaines, Illinois 60018, U.S.A. Martel Laboratories 7100 30th Avenue North St. Petersburg, Florida 33710, U.S.A.U.S. Geological Survey EROS Data Center Sioux Falls, South Dakota 57198, U.S.A. |
D. APPLICATIONS OF REMOTE SENSING TECHNOLOGY TO NATURAL HAZARD ASSESSMENTS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
1. Floods2. Hurricanes3. Earthquakes4. Volcanic Eruptions and Related Hazards5. Landslides6. Desertification
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
For purposes of assessing natural hazards in the context of integrated development planning studies, it is not necessary to have real-time or near real-time remote sensing imagery. What is required is the ability to define areas of potential exposure to natural hazards by identifying their occurrence or conditions under which they are likely to occur and to identify mechanisms to prevent or mitigate the effects of those hazards. This section considers the practical detectability by remote sensing technology of the potential for floods, hurricanes, earthquakes, volcanic eruptions and related hazards, and landslides. It will become evident that some of these hazards are interrelated, e.g., floods and hurricanes; earthquakes, volcanoes and landslides.
The ability to identify these natural hazards or their potential for occurring depends on the resolution of the image, the acquisition scale of the sensor data, the working scale, scenes free of clouds and heavy haze, and adequate textural and tonal or color contrast. The availability of stereomodels of the scene being studied can greatly enhance the interpretation. Figure 4-7 displays satellite remote sensing attributes to be taken into consideration for the assessment of various hazards.
After a hazard is identified, formulating appropriate mitigation measures and developing response plans may require different remote sensing data sets. This additional remote sensing data needed are most likely to include greater detail of the infrastructure, e.g., roads and facilities. This may have to be derived from aerial photography.
1. Floods
Floods are the most common of natural hazards that can affect people, infrastructure, and the natural environment. They can occur in many ways and in many environments. Riverine floods, the most prevalent, are due to heavy, prolonged rainfall, rapid snowmelt in upstream watersheds, or the regular spring thaw. Other floods are caused by extremely heavy rainfall occurring over a short period in relatively flat terrain, the backup of estuaries due to high tides coinciding with storm surges, dam failures, dam overtopping due to landslides into a reservoir, and seiche and wind tide effects in large lakes. Occasionally an eruption on a glacier or snow-covered volcanic peak can cause a flood or a mudflow in which the terrain is radically changed and any agrarian development is totally destroyed, frequently with much loss of life. See Chapter 8 for a more detailed discussion of flood hazards and Chapter 11 for a discussion of floods and mudflows associated with volcanic eruptions.
It is impossible to define the entire flood potential in a given area. However, given the best remote sensing data for the situation and a competent interpreter, the evidence for potential flood situations can be found or inferred. The most obvious evidence of a major flood potential, outside of historical evidence, is identification of floodplain or flood-prone areas which are generally recognizable on remote sensing imagery. The most valuable application of remote sensing to flood hazard assessments, then, is in the mapping of areas susceptible to flooding.
Synoptic satellite sensor coverage of a planning study area is the practical alternative to aerial photography because of cost and time factors. The application of Landsat MSS imagery to floodplain or flood-prone area delineation has already been demonstrated by comparing pre-flood scenes with scenes obtained at the height of the flood, using Landsat MSS band 7 (near IR) images in a color additive viewer (Deutsch et al., 1973). This temporal comparison can now be done pixel by pixel in a computer. Landsat TM, with its greater spatial resolution than MSS data (30m versus 80m) and its additional spectral coverage (7 bands versus 4 bands), can be used for more detailed mapping of floodplains and flood-prone areas on larger scale maps of 1:50,000 or greater. TM data have been used for discriminating land cover classification (Kerber et al., 1985) and to provide useful input to flood forecasting and flood damage models for urban and agricultural areas (Gervin et al., 1985).
Figure 4-7 - SATELLITE IMAGERY APPLIED TO NATURAL HAZARD ASSESSMENTS
EARTHQUAKES
|
VOLCANIC ERUPTION
|
LANDSLIDES
|
TSUNAMI
|
DESERTIFICATION
|
FLOODS
|
HURRICANES
| |
INFORMATION TO BE OBTAINED | Land-use maps, geological maps | Maps of areas vulnerable to lava flows, ash fall, debris fall and fires | Slope maps, slopes stability, elevation, geology, soil type, areas of standing water, land-use maps | Bathymetric/ topographic maps | Land-use maps, soil moisture content, crop condition and natural vegetation | Floodplain delineation maps, land-use classification, historical data, soil cover and soil moisture | Land-use maps |
SPECTRAL BAND | Visible and near IR | Visible, near IR and thermal IR | Visible | Visible, including blue and -near IR | Visible, near IR, and micro wave | Near IR, thermal IR and microwave | Visible and near IR |
SPATIAL RESOLUTION |
20-80m
|
30-80m
|
10-30m
|
30m
|
80m-1 km
| 20m (for cultural features); 30-80m (for land use); 1 km (for snow cover and soil moisture) | 20m (for cultural features); 30-80m (for land use) |
AREAL COVERAGE | Large area | Large area | Large area | Large coastal area | Large regional area | Large regional area | Large area |
ALL WEATHER CAPABILITY | No | No | No | No | No | No | No |
SYNOPTIC VIEW | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
STEREO CAPABILITY | Yes | Yes | Yes | Yes | No | Yes | No |
FREQUENCY OF OBSERVATIONS FOR PLANNING STUDY USE | 1 to 5 years | 1 to 5 years | 1 to 5 years | 1 to 5 years | Monthly | Seasonal (except weekly for snow cover and soil moisture) | Yearly |
Source: Adapted from Richards, PUB. The Utility of Landsat-D and other Satellite Imaging Systems in Disaster Management (Washington, DC: Naval Research Laboratory, 1986).
This approach to floodplain delineation does have its limitations. The area of potential flooding delineated in this manner may represent a flood level that exceeds an acceptable degree of loss. Also, no floods may have occurred during the period of the sensor operation. In this case, indirect indicators of flood susceptibility are used. A more detailed discussion of flood susceptibility and the use of Landsat imagery can be found in Chapter 8. Landsat and presumably similar satellite data floodplain indicators are listed in Figure 4-8.
There are large parts of tropical humid ecosystems where adequate Landsat or other similar imagery is not available due to cloud coverage or heavy haze. In some instances the heavy tropical vegetation masks many geomorphic features so obvious in drier climates. In this case the use of available radar imagery from space or previously acquired from an aircraft survey is desirable. The radar imagery, which has a resolution comparable to Landsat TM and SPOT from both space and sub-orbital altitudes, can satisfactorily penetrate the clouded sky and define many floodplain features. Moisture on the ground noticeably affects the radar return and, together with the textural variations emphasized by the sensor, makes radar a potentially desirable tool for flood and floodplain mapping.
2. Hurricanes
To mitigate the impact of hurricanes, the planner needs to know the frequency of storms of given intensity in the study area, to what extent these storms could affect people and structures, and what sub-areas would be most affected such as low-lying coastal, estuarine, and reverie areas threatened by flooding and storm surge. See Chapter 12 for a more detailed discussion of hurricanes and coastal areas.
The determination of past hurricane paths for the region can be derived from remote sensing data from the U.S. National Oceanographic and Atmospheric Administration (NOAA) satellite sensors designed and operated for meteorological purposes. These data are already plotted by meteorological organizations in the U.S.A. and other countries where hurricanes are a threat. For plotting new data, the best sensor is the AVHRR which, with its 2,700km swath, makes coverage twice a day, and has appropriate resolution. The red band is useful for defining daytime clouds and vegetation, while the thermal IR band (10.50 mm to 11.50 mm) is useful for both daytime and nighttime cloud observations.
Figure 4-8 LANDSAT FLOODPLAIN INDICATORS
- Upland physiography- Watershed characteristics, such as shape, drainage, and density- Degree of abandonment of natural levees- Occurrence of stabilized sand dunes on river terraces- Channel configuration and fluvial geomorphic characteristics- Backswamp areas- Soil-moisture availability (also a short term indicator of flood susceptibility)- Variations in soil characteristics- Variations in vegetation characteristics- Land use boundaries- Flood alleviation measures for agricultural development on the floodplain
Source: Adapted from Rango, A. and Anderson, A.T. "Flood Hazard Studies in the Mississippi River Basin Using Remote Sensing" in Water Resources Bulletin, vol. 10, 1974.
The AVHRR is not useful in other aspects of hurricane contingency planning due to its limited spatial resolution. These planning needs require higher resolution available from other satellite sensors. If imagery of areas inundated by floods, hurricane storms, or other storms is obtained with any sensor immediately after the event, it should be used, of course, regardless of its resolution. Any such information that is obtained in a timely fashion should be used to delineate the problem areas since their definition is more exact than can be interpreted from higher resolution data obtained during a non-flood period.
Predicting areas of potential inundation along coasts and inland can be achieved using topographic maps with scales as large as 1:12,500. When such maps are not available, remote sensing techniques can be used. In areas with a distinct wet and dry season, it is desirable to obtain information for the wet season from Landsat or comparable imagery in the near IR bands, or use a color IR composite made from Landsat MSS or TM imagery or from SPOT HRV imagery. These image products can be used to identify the moisture-saturated areas susceptible to flooding as well as the higher and drier ground for potential evacuation areas. Likewise, consideration of development plans in view of this potential natural hazard can proceed in a way similar to that for areas prone to flood hazards. For flood hazard assessments, radar imagery from space or aircraft could be used (if available) in lieu of the Landsat MSS imagery. Since there is a general lack of relief in low-lying coastal areas and estuarine areas, stereoscopy would not normally play an important role in this situation. However, stereoscopic viewing even without significant relief enhancement can still reinforce the details of the scene, although at considerably greater cost.
The development planner also needs to consider the additional feature of a hurricane-its high winds. In identifying measures to mitigate wind effects, the planner may consider the type of crops grown, if an agricultural development is being planned, and the design and construction materials used in buildings.
3. Earthquakes
The planning of development in earthquake-prone areas is laden with problems. There are large human settlements already located in earthquake/prone areas.
As with other geologic hazards, the frequency of occurrence can fall in cycles of decades or centuries. Earthquakes are particularly difficult to predict at this time. Thus, mitigation emphasis is on land use planning (non-intensive uses in most hazardous areas), on building strength and integrity, on response planning, and on incorporating mitigation measures into reconstruction efforts. The main problem is the identification of the earthquake damage-prone zones (see Chapter 11 for detailed discussion of earthquakes and their assessment). While in most areas of great earthquake activity some seismic information is available, it may not be sufficient for planning purposes. Remote sensing techniques and resulting data interpretation can play a role in providing additional information.
Tectonic activity is the main cause of destructive earthquakes, followed by earthquakes associated with volcanic activity. Where the history of earthquakes due to seismic activity is present in an area, the faults associated with the activity can frequently be identified on satellite imagery. Where volcanic-related earthquakes occur, the source is generally not as obvious: it may be due to movement on a fault near the surface or deep within the earth, to caldera collapse, or to magma movement within the volcanic conduit.
In order to identify earthquake hazards it is necessary to have the expertise to recognize them and then determine the correct remote sensing tools to best delimit them. Landsat imagery has been effectively and widely used for this purpose since it is less expensive and more readily available than other remote sensing data. Airborne radar mosaics have been successfully used for the delineation of fault zones. Generally, two mosaics can be made of an area: one with the far range portion of the SLAR and the other with the near range portion. The former is best used in areas of low relief where the relief needs to be enhanced, and the latter in areas of high relief where the shadow effect is not needed or may be detrimental to the image.
Radar is applicable to delineate unconsolidated deposits sitting on fault zones-upon which most of the destruction occurs-to identify areas where an earthquake can trigger landslides. This is best accomplished on stereomodels using adjoining and overlapping radar flight lines. Conventional aerial photography, in black and white or color, would also work well for this purpose.
An alternative, which is adequate but not as good as using radar and aerial photography, is to use multispectral imagery obtained from the Landsat TM and/or MSS or SPOT HRV sensors. Color IR composites or straight near-IR imagery from these sensors at scales up to about 1:100,000 can be used to define active surface fault zones, but not as efficiently as with the radar images. The distinction between bedrock versus unconsolidated materials and the areas of potential landslide hazards can be defined but, again, only if stereocoverage is available. SPOT sensors can provide this capability.
While radar imagery is an ideal data source, available coverage is extremely limited, and contracting airborne radar is usually prohibitively expensive. Landsat TM and MSS are the most practical data source, simply because of its availability, and both provide sufficient resolution for regional planning studies.
4. Volcanic Eruptions and Related Hazards
Many hazards are associated with the conditions brought about by volcanic activity. Active volcanoes pose hazards which include the immediate release of expelled ash, lava, pyroclastic flows, and/or poisonous hot gases; volcanic earthquakes; and the danger of mudflows and floods resulting from the rapid melting of snow and ice surrounding the vent during eruption. Some secondary hazards may threaten during volcanic activity or during periods of dormancy. These include landslides due to unstable accumulations of tephra, which may be triggered by heavy rains or by earthquakes. A more detailed discussion of volcanic hazards and their assessment can be found in Chapter 11.
Each volcano has its own particular behavior within a framework of given magmatic and tectonic settings. Prediction of a volcano's behavior is extremely difficult, and the best evidence for the frequency of activity and its severity is the recorded history of eruptions. Imminent eruptions are now best recognized by on-site seismic monitoring. Some classifications distinguish between active, inactive, dormant, and extinct volcanoes. But since some of the most catastrophic eruptions have come from "extinct" volcanoes, many volcanologists have abandoned such a classification, settling for a simple distinction between short-term and long-term periodicity.
Gawarecki et al. (1965) first detected volcanic heat from satellite remote sensing using thermal IR imagery from the high resolution IR radiometer (HRIR). Remote sensing data interpretation can lead to the recognition of past catastrophic events associated with recently active volcanoes (recently in the geologic sense), as in the Andes and the Lesser Antilles. This information together with the available historical data can be used as the basis of assessing the risks of an area with potential volcano-related hazards.
The varied nature and sizes of volcanic hazards require the use of various types of sensors from both satellites and aircraft. The relatively small area involved with volcanoes should encourage the use of aerial photography in their analysis. Panchromatic black and white stereo aerial coverage at scales between 1:25,000 to 1:60,000 is usually adequate to recognize and map geomorphic evidence of recent activity and associated hazards. Color and color IR photography may be useful in determining the possible effects of volcanic activity on nearby vegetation, but the slower film speed, lower resolution, and high cost diminish much of any advantage they provide.
The airborne thermal IR scanner is probably the most valuable tool in surveying the geothermal state of a volcano. The heat within a volcano and underlying it and its movement are amenable to detection. Because of the rapid decrease in resolution with increasing altitude (about 2m per 1,000m) the surveys need to be made at low altitudes under 2,000m.
An IR pattern of geothermal heat in the vicinity of a volcano is an indication of thermal activity which many inactive volcanoes display. Many volcanoes thought to be extinct may have to be reclassified if aerial IR surveys discover any abnormally high IR emissions from either the summit craters or the flanks. Changes in thermal patterns can be obtained for a volcano only through periodic aerial IR surveys taken under similar conditions of data acquisition. The temperature and gas emission changes, however, can be monitored on the ground at ideal locations identified on the thermal imagery, making periodic overflights unnecessary. Continuous electronic monitoring of these stations is possible by transmission through a geostationary data relay satellite, another phase of remote sensing.
The thermal IR bands of the satellite sensors now available have inadequate spatial and thermal resolution to be of any significant value to detect the dynamic change in volcanic geothermal activity. In addition to sensing geothermal heat, however, other remote sensing techniques are useful in preparing volcanic hazard zonation maps and in mitigating volcanic hazards. Mitigation techniques requiring photo interpretation and topographic maps include predicting the path of potential mudflows or lava flows and restricting development in those areas.
5. Landslides
Landslides, or mass movements of rock and unconsolidated materials such as soil, mud, and volcanic debris, are much more common than is generally perceived by the public. Many are aware of the catastrophic landslides, but few are aware that small slides are of continuous concern to those involved in the design and construction business. These professionals can often exacerbate the problem of landsliding through poor planning, design, or construction practices. Frequently, the engineer and builder are also forced into difficult construction or development situations as a result of ignoring the potential landslide hazard. This can be avoided if there is early recognition of the hazard and there is effective consultation between planners and the construction team prior to detailed development planning. See Chapter 10 for a more detailed discussion of landslide hazards and their assessment.
The mass movement of bed rock and unconsolidated materials results in different types of slides, magnitudes, and rates of movement. An area with a potential landslide hazard usually has some evidence of previous occurrences, if not some historical record. Unfortunately, some types of landslides, particularly those of small size, cannot be delineated on remote sensing imagery or through aerial photography. Usually the scars of the larger slides are evident and, although the smaller slide features may not be individually discerned, the overall rough appearance of a particular slope can suggest that mass movement occurred. If a fairly accurate geologic map is available at a reasonable scale (1:50,000 or larger) rock types and/or formations susceptible to landslides may be examined for evidence of movement. An example of this would be finding a shale in a steeper than usual slope environment, implying the strong possibility of a landslide history. An examination of stream traces frequently shows deflections of the bed course due to landslides. If one can separate out the tectonically controlled stream segments, those deflections due to slides or slumps often become evident.
Typical features that signify the occurrence of landslides include chaotic blocks of bedrock whose only source appears to be upslope; crescentic scarps or scars whose horns point downward on a normal-looking slope; abnormal bulges with disturbed vegetation at the base of the slope; large intact beds of competent sedimentary or other layered rock displaced down dip with no obvious tectonic relationship; and mudflow tongues stretching outward from the base of an obviously eroded scar of relatively unconsolidated material. A good understanding of the structural geology of the study area frequently places these superficial anomalies into perspective. As discussed in Chapter 10, the susceptibility to landslides is relative to the area. Landslides can occur on gentle slopes as well as on steep slopes, depending on landscape characteristics.
Most landslide discussions do not address the problem of sinkholes, which are a form of circular collapse landslides. The karstic areas in which they occur are easy to identify even on some satellite imagery (MSS, TM, SPOT, etc.) due to their pitted appearance and evidence of the essentially internal drainage. Despite the obvious occurrence of many sinkholes, many individual small sinkholes are subtle and not easily recognized. These are frequently the sites of collapse and subsequent damage to any overlying structure when ground water is removed to satisfy development needs, which results in lowering the water table and undermining the stability of the land.
The spatial resolution required for the recognition of most large landslide features is about 10m (Richards, 1982). However, the recognition depends to a great extent on the ability and experience of the interpreter and is enhanced by the availability of stereoscopic coverage, which can be expensive to acquire. Stereoscopic coverage and the resolution requirements preclude use of most satellite-borne sensor imagery, although large block landslides can be detected on Landsat MSS and TM imagery.
Given the spatial resolution requirement, SPOT HRV-P (panchromatic mode) imagery can be useful with its 10m resolution. Its broad band coverage, however, is not conducive to providing adequate contrast in scenes involving heavily vegetated tropics, where most of the potential hazards occur. Ameliorating this factor slightly is the availability of stereocoverage. It is important to understand that this is specifically programmed for the SPOT satellite and that stereocoverage is not normally acquired during sensor operation.
Detection of landslide features is more easily achieved using airborne sensors. Aerial photography with its normal stereoscopic coverage is the best sensor system with which to define landslides, both large and small. Aerial photographic scales as small as 1:60,000 can be used. Black and white panchromatic or IR films are adequate in most cases, but color IR may prove better in some instances. The IR-sensitive emulsions, as stated earlier, eliminate much of the haze found in the humid tropics. The open water or other moisture in back of recent slump features stands out as an anomaly in the aerial IR stereomodel, either in black and white or color. The color IR photography might, in some rare cases, show the stress on the vegetation caused by recent movement. If the scales are large enough, tree deformation caused by progressive tilting of the slope of the soil might also be detected.
A more sensitive detector of moisture associated with landslides is the thermal IR scanner. This sensor is particularly useful in locating seepage areas that lubricate slides. It is particularly effective during the night, when there is a maximum temperature difference between the terrain and the effluent ground water. Despite its utility many factors rule out the widespread use of the thermal IR scanner. These factors include the low altitude required for reasonable spatial resolution, the large number of flight lines required for the large area involved, and the geometrical distortions Inherent in the system. If the terrain to be interpreted has some relief and is nondescript, these distortions become an even greater problem when the data are interpreted by making the location of features very difficult.
SLAR, especially the X-band synthetic aperture radar with its nominal 10m resolution, can be marginally useful in a stereo mode because of its ability to define some larger textures related to landslides. In some cloud-prone environments radar may be the only sensor that can provide interpretable information.
6. Desertification
Desertification occurs when an ecosystem experiences a diminution or loss of productivity. This process can have a natural and an anthropic component, which may reinforce each other, creating a synergetic effect (see Chapter 9). The degree of desertification risk is directly related to certain natural conditions such as climate, topography, natural vegetation, soil, and hydrology, as well as to the intensity and type of anthropic activity in the area. Desertification is among the most serious problems of the region. This trend indicates the increasing need to consider desertification processes in integrated development planning studies. Remote sensing, both spaceborne and airborne, provides valuable tools for evaluating areas subject to desertification. Film transparencies, photographs, and digital data can be used for the purpose of locating, assessing, and monitoring deterioration of natural conditions in a given area. Information about these conditions can be obtained from direct measurements or inferred from indicators (keys to the recognition of a desertification process).
In order to describe, evaluate, and decide about the type of action to be taken, the following issues should be addressed:
- Location: involves the identification of areas that are currently undergoing desertification and areas expected to be exposed to the forces that can lead to deterioration.- Assessment: involves the identification and quantification of vegetative cover types, soils, land forms, and land-use change patterns. Vulnerability to change, rate of change, and direction of change in desertification patterns can be studied through this assessment.- Monitoring: accomplished by detecting and measuring changes in landscape characteristics over a period of time. Comparisons are made between present conditions and previously observed conditions for the purpose of recording the reduction in biological productivity.
Chapter 9 presents an initial assessment technique utilizing information commonly available in the early integrated development planning stages. For a more detailed approach, four sets of data should be taken into consideration for a desertification study of a given area: a set taken at the end of the humid season, a set taken at the end of the last dry season, and the same two seasons taken five or ten years earlier (López Ocaña, 1989). Data selection for a given area will be directly related to the desired amount of detail, size of the area, required degree of precision and accuracy, and available time frame.
Large-scale aerial photography provides a great amount of detail for this type of study. Systematic reconnaissance flights can be used for environmental monitoring and resource assessment. Radar sensors and infrared scanners may be used to monitor soil moisture and other desertification indicators. However, acquisition of this type of data is costly and time consuming.
The use of satellite imagery is recommended during the first stages of a detailed desertification study since it offers an overview of the entire region. Computer enhancements, false color composites, and classifications can offer useful information. Optical enhancements can be performed, but these lack the quantitative control available through an automated approach. Statistical data obtained from a quantitative analysis through the use of a geographic information system (GIS-see Chapter 5) can be expressed as a histogram, a graph, a table, or a new image.
AVHRR imagery is commercially available and has been used for vegetation change studies. Ground resolution of 1 to 4 km represents some limitation in making large continental area studies. Other studies have used Nimbus data to delineate moisture patterns and vegetation boundaries. Geostationary Operational Environmental Satellite (GOES) data have been used effectively to locate and measure dust plumes; and Seasat SAP imagery has been applied in the delineation of large dune morphology.
Landsat MSS and TM and SPOT data have proven to be useful and cost effective for regional assessments. Landsat transparencies of bands 5 and 7 have been used to monitor superficial changes in areas undergoing desertification, and to map present water bodies and former drainage systems. Temporal tonal variations on Landsat MSS have been correlated with variations on the field. Movement of sand-dune belts has been detected using Landsat with a multitemporal approach. Albedo changes in arid terrains have been calculated using Landsat digital data: phenomena that tend to lower productivity (increased erosion, loss of vegetation density, deposition of colic sedimentation) also tend to appear brighter on the image. On the contrary, phenomena that tend to increase productivity (increased vegetation, soil moisture), tend to darken the land. In this way, brightness variations can be detected in an area over a period of time. These data can also be calibrated with ground data collected from the areas where change has occurred.
Aerial and space remote sensing provide valuable tools for desertification studies, although, as for any other natural hazard related study, they must be combined with ground-collected data. The use of remote sensing methods should minimize the need for ground data, therefore saving time and resulting quite inexpensive per unit of data. The combination of remotely sensed and ground-collected data can then provide the basis for the assessment.
REFERENCES
American Society of Photogrammetry (ASP). Manual of Remote Sensing, 2nd ed. (Falls Church, Virginia: ASP, 1983).
Bertaud, M.A. The Use of Satellite Images for Urban Planning. A Case Study from Karachi, Pakistan. The World Bank Technical Note (Washington, D.C.: The World Bank, 1989).
Budge, T. A Directory of Major Sensors and Their Parameters (Albuquerque, New Mexico: Technology Application Center, University of New Mexico, 1988).
Carter, D., et al. "Space Applications for Disaster Mitigation and Management" in Acta Astronautica, Vol. 19, No. 3 (Great Britain: Pergamon Press, 1989), pp. 229-249.
Deutsch, M., et al. "Mapping of the 1973 Mississippi River Floods from the Earth Resources Satellite (ERTS)" in Proceedings No. 17 - Remote Sensing and Water Resources Management (Bethesda, Maryland: American Water Resources Association, 1973), pp. 39-54.
Deutsch, M., et al. "Quick Response Monitoring of Flood Disasters from Satellite Imagery" in Proceedings of the Twentieth International Symposium on Remote Sensing of the Environment (Ann Arbor, Michigan, 1986).
Gawarecki, S.J., et al. "Infrared Spectral Returns and Imagery of the Earth from Space and Their Applications to Geologic Problems" in Scientific Experiments for Manned Orbital Flight. Science and Technology Series, American Astronautical Society. Vol. 4 (1965), pp. 13-33.
Gawarecki, S.J., Moxham, P.M., Morgan, J.Q., and Parker, D.C. "An Infrared Survey of Irazu Volcano and Vicinity, Costa Rica" in Proceedings of the 14th International Symposium on Remote Sensing of the Environment (San Jose, Costa Rica, April 1980), pp. 1901-1912.
Gervin, J.C., et al. 'The Effect of Thematic Mapper Spectral Properties on Land Cover Mapping for Hydrologic Modeling" in Proceedings of the U.S. Army Corps of Engineers Fifth Remote Sensing Symposium (Ft. Belvoir, Virginia: Water Resources Support Center, 1985), pp. 249-260.
Hassan, H., and Luscombe, W. "Disaster Information and Technology Transfer in Developing Countries" in Proceedings of the Colloquium on the Environment and Natural Disaster Management (Washington, D.C.: The World Bank, 1990).
Kerber, A.G., et al. "Floodplain Land Cover Mapping Using Thematic Mapper Data" in Proceedings of the U.S. Army Corps of Engineers Fifth Remote Sensing Symposium (Ft. Belvoir, Virginia: Water Resources Support Center, 1985), pp. 262-271.
Kruus, J.M., et at. "Flood Applications of Satellite Imagery" in Deutsch, M., Wiesnet, D.R., and Rango, A.R. (eds.), Satellite Hydrology (Bethesda, Maryland: American Water Resources Association, 1981), pp. 292-301.
Lillesand, T.M., and Kiefer, R.W. Remote Sensing and Image Interpretation (Somerset, New Jersey: John Wiley and Sons, Inc., 1987).
Lopez Ocaña, C. Desertification Risks Assessment in Development Planning. Unpublished manuscript (Washington, D.C.: World Resources Institute, 1989).
Morgan, G. Satellite Remote Sensing Technology for Natural Hazards Preparedness and Emergency Response Planning. World Bank, Environment Operation and Strategy Division (Washington, D.C.: World Bank, May 1989).
Nossin, J. "Aerospace Survey of Natural Hazards: the New Possibilities" in The International Institute for Aerospace Survey and Earth Sciences (ITC) Journal, 1989-3/4 (Enschede, The Netherlands: ITC, 1989).
Rango, A., and Anderson, A.T. "Flood Hazard Studies in the Mississippi River Basin Using Remote Sensing" in Water Resources Bulletin, Vol. 10, No. 5(1974), pp. 1060-1081.
Organization of American States. Physical Resource Investigations for Economic Development (Washington, D.C.: Organization of American States, 1969).
Richards, P. B. The Utility of Landsat-D and Other Satellite Imaging Systems in Disaster Management, Final Report. NASA Goddard Space Flight Center Disaster Management Workshop, NASA DPR S-70677 (Washington, D.C.: Naval Research Laboratory, March 29-30, 1982).
Sabins, F. F., Jr. Remote Sensing: Principles and Interpretation (New York: W.H. Freeman, 1986).
Schaber, G.G., et al. "Shuttle Imaging Radar: Physical Controls on Signal Penetration and Subsurface Scattering in the Eastern Sahara" in IEEE Trans. Geoscience Remote Sensing, Vol. GE-24 (1986), pp. 603-623.
Schroeder, M. "Spacelab Metric Camera Experiments" in Satellite Remote Sensing for Resources Development. (Gaithersburg, Maryland: Graham and Trotman Ltd., 1986), pp. 81-92.
Schuster, R.L, and Krizek, R.J. (eds.). Landslides: Analysis and Control. Transportation Research Board Special Report 176 (Washington, D.C.: National Academy of Sciences, 1978).
Sellers, S.C., Rango, A., and Henninger, D.L. "Selecting Reconnaissance Strategies for Floodplain Surveys" in Water Resources Bulletin, Vol. 14, No. 2 (1978), pp. 359-373.
Water, L. "Uses of Satellite Technology in Disaster Management" in Communication When It's Needed the Most: How New Technology Could Help in Sudden Disasters. Report of the International Disaster Communications Project (Washington, D.C.: The Annenberg Washington Program, 1990).
Weber, C. "Remote Sensing and Natural Hazards. Contribution of Spatial Imagery to the Evaluation and Mitigation of Geological Hazards" in Proceedings of the 27th International Geological Congress, Vol. 18 (Moscow: VNU Science Press, 1984), pp. 211-228.
Wiesnet, D.R., and Deutsch, M. "Flood Monitoring in South America from the Landsat, NOAA and Nimbus Satellites" in XXVI COSPAR 86 (Toulouse, France, 1986).
Wiesnet, D.R., Scott, R.B., and Matson, M. "The NOAA Satellites: A Largely Neglected Tool in the Land Sciences" in XXIV COSPAR 82 (Ottawa, 1982).
Zimmerman, P. 'The Role of Remote Sensing in Disaster Relief in Communication When It's Needed Most: How New Technology Could Help in Sudden Disasters. Report of the International Disaster Communications Project (Washington, D.C.: The Annenberg Washington Program, 1990).
Full Text
PREFACE
CHAPTER 1 - INCORPORATING NATURAL HAZARD MANAGEMENT INTO THE DEVELOPMENT PLANNING PROCESS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
A. WHAT ARE NATURAL HAZARDS?
B. SUSCEPTIBILITY TO VULNERABILITY REDUCTION
C. HAZARD MANAGEMENT AND DEVELOPMENT PLANNING
D. HAZARD MANAGEMENT IN SELECTED ECONOMIC SECTORS
E. IMPLEMENTING THE RECOMMENDATIONS: STRATEGIES FOR DEVELOPMENT ASSISTANCE AGENCIES1
REFERENCES
CHAPTER 2 - NATURAL HAZARD RISK REDUCTION IN PROJECT FORMULATION AND EVALUATION
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
A. NATURAL HAZARDS IN PERSPECTIVE
B. BASIC CONCEPTS: NATURAL HAZARDS AND INVESTMENT PROJECTS
C. THE USE OF NATURAL HAZARD INFORMATION IN INVESTMENT PROJECT PREPARATION
D. INCORPORATING NATURAL HAZARDS INTO PLANNING AND DECISION-MAKING IN THE PUBLIC SECTOR
E. PRINCIPLES OF ECONOMIC ANALYSIS
F. INCORPORATING NATURAL HAZARDS INTO THE ECONOMIC ANALYSIS OF INVESTMENT PROJECTS
G. CONCLUDING REMARKS
REFERENCES
CHAPTER 3 - RESOURCE EVALUATION AND THE ROLE OF ECOSYSTEMS IN MITIGATING NATURAL HAZARDS
A. LAND-USE EVALUATIONS IN LATIN AMERICAB. LIMITATIONS OF LAND-USE EVALUATIONS
C. LAND-USE EVALUATIONS BASED ON A SYSTEMS VIEW
D. ASSESSING NATURAL HAZARDS IN LAND-USE EVALUATIONS
E. NATURAL SERVICES IN SUPPORT OF HAZARD MITIGATION
REFERENCES
CHAPTER 4 - REMOTE SENSING IN NATURAL HAZARD ASSESSMENTS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 5 - GEOGRAPHIC INFORMATION SYSTEMS IN NATURAL HAZARD MANAGEMENT
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 6 - MULTIPLE HAZARD MAPPING
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 7 - CRITICAL FACILITIES MAPPING
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 8 - FLOODPLAIN DEFINITION AND FLOOD HAZARD ASSESSMENT
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 10 - LANDSLIDE HAZARD ASSESSMENT
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 11 - GEOLOGIC HAZARDS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
CHAPTER 12 - HURRICANE HAZARDS
![](https://www.oas.org/dsd/_themes/tabs/atabrule.gif)
ليست هناك تعليقات:
إرسال تعليق