 |
Remote SensingRelated concepts: [ Resolution | Ground Stations | Orbits | Spectral Coverage | Scene Size | Stereo Pairs | Synoptic Views | Turnaround Time ] Other Remote Sensing Resources on the Web Interested in the design of the first U.S. commercial radar satellite? Give us your feedback. Remote sensing generally refers to taking images of objects on or near the Earth's surface, by means of observations from airplanes or satellites. Users of remote sensing find that it is the most economical means for viewing what they need to see, when they need to see it. Both airborne and satellite-based remote sensing systems view the Earth from overhead, with airplanes generally providing very high resolution over limited areas, and satellites providing lower resolution but over the entire planet. Often, satellite-based remote sensing is the only way to see what is going on in a particular part of the world, e.g., one that is inaccessible to overflight due to political or military restrictions. A remote sensing satellite system entails one or more orbiting spacecraft with a stable platform for sensors (cameras) and a means of transmitting data back to the ground. Most satellite images are digital, although a few older systems, such as the Soviet Resurs-F system, still generate and deliver imagery on photographic film. The bulk of aerial remote sensing is still photographic.
The figure illustrates how, as the resolution of an image increases, the area imaged on the ground by each pixel decreases and the number of pixels increases. For a tenfold decrease in the side length of a pixel (e.g., from 30 meters to 3 meters), the number of pixels increases one hundredfold, in other words, there is one hundred times as much information provided about a given area.
First, they provide a means for receiving health-and-status reports from the satellite, uploading commands to image particular regions of the planet, and uploading new software. These functions are normally referred to as TT&C, for Telemetry, Tracking and Control. The second function for ground stations is to receive images downlinked from the satellite. Ground stations are often located at high latitudes to benefit from being contacted nearly once every orbit, thus minimizing the number of stations required to contact the satellite at any given time. High latitudes can be quite inhospitable, however, driving up the operation cost of the ground station (particularly if it is located far away from utilities and communications links).
While there are a variety of possible orbit types, a remote sensing system is interested primarily in circular orbits, that is, orbits whose path is a circle centered on the center of the Earth. The shape of these orbits is determined by two key parameters: the orbital altitude, or the height above the surface of the Earth, and the inclination, or the angle made between the circle of the orbit and the equator.
For any given altitude, there is a precise inclination that will result in a sun-synchronous orbit. A satellite in such an orbit always passes directly over a given spot on the ground at the same local time, in other words, with the sun in essentially the same position in the sky from one day to the next. As a result, shadows in pictures taken on different days are at the same angles, making it easy to detect changes over time. Furthermore, the sun is always in the same position relative to the satellite, simplifying the design of the satellite's solar power system and reducing its cost.
Other users require color to determine the types of objects, such as vegetation, soil, water, or clouds. These users are said to need multispectral coverage. The colors provided by a remote sensing system need not be limited to visible light, but may include infrared radiation as well. There are many possible bands that can be used by a remote sensing system. Some are ideal for one set of applications, while others are good for a different set. Since the technology for providing a large number of different bands is currently in the R&D phase and immature, builders of a remote sensing system must make compromises between many potential users to choose the set of bands that will satisfy the largest number of users.
Both high resolution and large scene size drive up the number of pixels in an image. For example, a full Landsat Thematic Mapper scene of 185 x 170 km is roughly 35,000,000 pixels, each of which has 30 meter resolution. The same 35,000,000 pixels would cover a scene only 18.5 x 17 km at 3 meter resolution. If the entire 185 x 170 km scene were viewed with 3 meter resolution, it would require 3,500,000,000 pixels! Since the scarce resources in building a satellite imager are communications bandwidth to the ground and on-board storage capacity, these tend to place an upper bound on the number of pixels that are practical in an image. While both high resolution and large scene size would be nice to have, they are normally traded off to avoid prohibitive satellite, communications, and processing costs. In most cases, the applications that need broad area views neither need nor are benefited by high resolution; mineral and oil exploration surveys appear to be adequately serviced by currently available satellite image resolution (10-30 meters). Applications such as mapping that do require high resolution over large areas can achieve their objectives by obtaining multiple high-resolution images of narrower areas and assembling them to form a single, large-area mosaic image. Thus, the small scene size of a single high resolution (1-3 meters) image is not a competitive disadvantage to those users who require large images; all they need to do is request multiple adjacent images.
There are two kinds of stereo pair: side-side and in-track. Side-side stereo pairs are two images taken of the same region on the ground taken on different passes of the satellite over that region. The first, for example, might be taken on a pass by the satellite to the right of the desired region, and the second taken several days later when the satellite passed either directly overhead or to the left of that region. In-track stereo pairs, in contrast, are taken by a satellite on the same pass over a given region, but at slightly different times. In-track stereo pairs are also known as fore-aft. The advantage of in-track stereo pairs over side-side is the near-simultaneity of image capture, which ensures that the two images are truly of the same scene as opposed to the same region on different days. Because of subtle variations in lighting, wind and atmospheric conditions, rainfall, snow cover, and other effects, the simultaneity of in-track stereo pairs makes them much more valuable for cartography and related applications than side-side stereo pairs.
>From an airplane at 3 km altitude (roughly 9750 feet), an object that is 3 km off to the side appears compressed by 200%. From a spacecraft at 400 km altitude, the same object that is 3 km off to the side hardly appears compressed -- only .0056%.
1) The time it takes for a requested image to be taken by a satellite and delivered to the ground (determined by the remote sensing system architecture, including satellite design, choice of orbit, and the number and location of ground stations). 2) The time it takes to process an image on the ground once it has been received from the satellite system and deliver it to the user (determined by the degree of automation in ground processing and the choice of data delivery mechanism). Timeliness requirements for imagery vary substantially. Media, intelligence, and disaster control/assessment efforts all require images as soon as possible. Turnaround times of several days or better are needed, but weeks-to-months turnaround is the norm to date. Long-term forestry inventory, mapping, and minerological and oil exploration have no real timeliness requirement; these groups often rely on archived data. Intermediate between these two applications, with a need for data within one to several days, are users of remote sensing for agricultural crop evaluation and forecasting. |
||||||||||||||||||||||||
 |
©1997 EarthWatch Incorporated. All rights reserved. |
