Abstract: In this term paper I would like to present an overview of application of signal processing in satellite imaging. There is no other technology that can compare to the versatility of satellite imaging with its immediate playback. Whether we are in management or in technical position this term paper will help to understand the immense role of signal processing in satellite imaging.
INTRODUCTION
SATELLITE IMAGING
Fig.1 Satellite imaging
The first crude image taken by the satellite Explorer 6 shows a sunlit area of the Central Pacific Ocean and its cloud cover. The photo was taken when the satellite was about 17,000 mi (27,000 km) above the surface of the earth on August 14, 1959. At the time, the satellite was crossing Mexico.
Satellite imaging consists of photographs of Earth or other planets made by means of artificial satellites.
HISTORY
Fig.2 The first television image of Earth from space transmitted by the TIROS-1 weather satellite.
The first satellite photographs of Earth were made on August 14, 1959 by the U.S. satellite Explorer 6. The first satellite photographs of the Moon might have been made on October 6, 1959 by the Soviet satellite Luna 3, on a mission to photograph the far side of the Moon. The Marble photograph was taken from space in 1972, and has become very popular in the media and among the public. Also in 1972 the United States started the Landsat program, the largest program for acquisition of imagery of Earth from space. Landsat 7, the most recent Landsat satellite, was launched in 1999. In 1977, the first real time satellite imagery was acquired by the USA's KH-11 satellite system.
All satellite images produced by NASA are published by Earth Observatory and are freely available to the public. Several other countries have satellite imaging programs, and a collaborative European effort launched the ERS and Envisat satellites carrying various sensors. There are also private companies that provide commercial satellite imagery. In the early 21st century satellite imagery became widely available when affordable, easy to use software with access to satellite imagery databases became offered by several companies and organizations.
APPLICATIONS OF SIGNAL PROCESSING IN SATELLITE IMAGING
Satellite Imaging systems were developed in the 1950s mainly by the armed forces. Radar is an active remote sensing system which means that it provides its own source of energy to produce an image. It therefore does not require sunlight (as do optical systems) and data can be acquired either by day or by night. Furthermore, due to the specific wavelength of radar, cloud cover can be penetrated without any effect on the imagery.
Synthetic Aperture Radar (SAR) is a technique for creating high resolution images of the earth's surface. Over the area of the surface being observed, these images represent the backscattered microwave energy, the characteristics of which depend on the properties of the surface, such as its slope, roughness, humidity, textural in homogeneities and dielectric constant. These dependencies allow SAR imagery to be used in conjunction with models of the scattering mechanism to measure various characteristics of the earth's surface, such as topography. SAR has become a valuable remote sensing tool for both military and civilian users. Military SAR applications include intelligence gathering, battlefield reconnaissance and weapons guidance. Civilian applications include topographic mapping, geology and mining, oil spill monitoring, sea ice monitoring, oceanography, agricultural classification and assessment, land use monitoring and planetary or celestial investigations.
The image shows a SAR radar map covering the whole of Germany
Another highly active research area in radar remote sensing is repeat pass satellite SAR interferometry (InSAR). InSAR provides a means for measuring displacements of the solid earth, glaciers, ice sheets, and fast sea ice to an accuracy of fractions of a radar wavelength (a few cm) during the time intervals between observations, using synthetic aperture radar (SAR) imagery. Since the launch of the first European Remote Sensing satellite (ERS-1) in 1991, this rapidly-evolving technology has been employed to measure, for example, co seismic displacements; the motion of glaciers and ice sheets in Alaska, Greenland, Antarctica and elsewhere; retreat of the grounding line of a major West Antarctic ice stream; deflation of a European volcano following an eruption; and crustal extension of potentially active volcanic vents in SW Alaska. In addition, InSAR can be employed to derive digital elevation models (DEMs) of the Earth's surface. Other applications of InSAR include prediction of earthquakes and volcanic eruptions, ice flow mapping, forest mapping and land classification. The limitations caused by atmospheric effects presently seem to be the most fundamental and severe limitation for this otherwise incredibly sensitive technique. Furthermore, the correlation map that used to be "just" a by-product of the interferometric processing, and at best a measure of the interferogram quality, is now becoming important information in itself. Correlation maps are used for volume scattering estimation and forest height measurement as well as for land use classification.
Interferogram
Radar Image
The scene shows an area near White Sands, New Mexico, and USA and covers approximately 50 km x 150 km.
The individual phase values appear as coloured rings. The steeper the slopes, the closer the fringes. Topography can already be seen directly in the interferogram.
Another application area in radar remote sensing is hydrology, including the retrieval of soil moisture and snow water content, glaciology, and radar mapping of vegetation. Hydrology is an area where SAR and also active imaging radar of lower resolution have much to offer. In relation to soil moisture estimation, polarimetric data have proven capabilities. Difficult problems include the vegetation cover and the requirement that the soil type/texture needs to be known. There is, however, hope that these problems can be mitigated. Using lower frequencies, e.g. P-band, enables penetration of low to moderate vegetation. More interestingly, the soil texture can potentially be estimated from a time series of measurements during a drying period following precipitation.
False colour intensity composite of two ERS passes over the Welkom goldfields.
The mapping of forest and biomass, as well as agricultural crops, is also active application areas. Many techniques show promise with respect to forest and biomass mapping and it has been shown that the backscatter coefficient of very low frequency systems (UHF and VHF) does not saturate at as small biomass values as the more common frequencies at L-band and C-band.
REMOTE SENSING APPLICATIONS BY INSTRUMENT
Wind Scatterometer (WSC) Applications
Wind scatterometers use accurate measurements of the radar backscatter from the ocean surface when illuminated by a microwave signal with a narrow spectral bandwidth to derive information on ocean surface wind velocity. At a given angle to the flight path of the satellite, the amount of backscatter depends on two factors, namely the size of the surface ripples of the ocean and their orientation with respect to the propagation direction of the pulse of radiation transmitted by the scatterometer. The first is dependent on wind stress and hence wind speed at the surface, while the second is related to wind direction.
This map displays the ocean surface winds at 10m on the 28th July 2000 from the ERS-2 scatterometer.
Scatterometer instruments aim to achieve high accuracy measurements of wind vectors, and resolution is of secondary importance. The resolution of the ERS scatterometer is 50 km, though the grid sampling is 25 km. Because the scatterometer operates at microwave wavelengths, the measurements are available irrespective of weather conditions. The assimilation of scatterometer data into atmospheric forecasting models greatly improves the description of cyclonic features so important in predicting future weather patterns. There are numerous other applications, such as the measurement of sea ice extent and concentration, and emerging land applications such as regional-scale monitoring of ice shelves, rainforests and deserts.
Radar Altimeter (RA) Applications
The radar altimeter is designed to make accurate measurements of the satellite's height above the sea surface which is then converted to the sea surface's height above a reference ellipsoid. When the altimeter takes a height measurement, it is measuring a height contributed to by many different types of phenomena, from the underlying marine geoid, through the large-scale general circulation of the oceans, to mesoscale eddies 100 km across. In addition to highly precise height measurements, the altimeter makes measurements of the heights of waves that appear in its footprint, and of surface wind speed.
Applications of the radar altimeter include:
Measuring the marine geoid.--Information has been extracted from altimeter data, particularly that provided by the high resolution dedicated Geodetic Mission of ERS-1, to provide maps of average sea surface topography - the marine geoid. The geoid is the fundamental reference surface of geodesy. Through its use in geoid determination, altimetry aids in revealing the location of ocean floor features such as faults, trenches, spreading zones, sea mounts and hot spots. Information may also be gained on the age, structure and dynamics of the lithosphere, particularly in the area of subduction zones, leading to a better understanding of the relationship between the lithosphere and the mantle, and of mantle convection. Additional, commercially valuable information can be derived on potential locations of oil-bearing structures using the effect that low density deposits (such as crude oil) have on the shape of the gravity field. This information has been derived not only over oceans, but also in the Arctic Ocean, using altimetry over sea ice.
Measuring sea state.--The radar altimeter also measures the heights of waves that appear within its `footprint', and the wind speed at the sea surface. Near real time measurements of Significant Wave Height (SWH) by the ERS altimeter are assimilated operationally into wave models to provide wave forecasts, essential for the optimisation of a range of marine operations.
Measuring the topography of the oceans.--Worldwide sea level varies significantly in space and time. Regional variations in sea level occur as a result of pressure differentials within the ocean, which result from momentum and heat flux exchange with the atmosphere. The resultant differences in sea level are thus directly related to ocean currents. Ocean topography can bemeasured directly and monitored for change using the ERS radar altimeter. Along with data from other similar instruments, the information can be assimilated into ocean circulation models which transform satellite surface information into three-dimensional descriptions of ocean currents and transports. An important fluctuation in the ocean-atmosphere system is the El Nino Southern Oscillation (ENSO) phenomenon, which causes an increase in ocean temperatures throughout the central and tropical Pacific which can produce dramatic changes in climate on the timescale of months to years. The events associated with ENSO can be measured in sea surface topography by the ERS altimeter, and in sea surface temperature by the ERS Along Track Scanning Radiometer (ATSR).
Along Track Scanning Radiometer (ATSR) Applications
Remote sensing data from the ERS-2 ATSR-2 allows the monitoring of agricultural fires and wildfire distribution on a global scale and in near real time. All hot spots (including gas flares) with a temperature higher than 312 K at night are precisely located (better that 1 km). Data from the ATSR sensor is also used for volcano monitoring applications and measuring ocean skin temperatures.
Global Ozone Monitoring Experiment (GOME) Applications
Atmospheric ozone and NO2 global monitoring have been going on since GOME products became available (July 1996). Additional applications could stem from on going scientific studies, as GOME data can also be used for retrieving other trace gases relevant to the ozone chemistry as well as other atmospheric constituents and climatic variables like clouds, aerosols and solar index, all of which are crucial for assessing climate change.
Synthetic Aperture Radar (SAR) Applications
Observations of the Earth using Synthetic Aperture Radar (SAR) have a wide range of practical applications, such as:
On the oceans:
Most of the man-made illegal or accidental spills are highly visible on radar images. Ships can be detected and tracked from their wakes. Natural seepage from oil deposits can also be observed, providing hints for the oil industries. Scientists are studying the radar backscatter from the ocean surface which is related to wind and current fronts, eddies, and internal waves. In shallow waters SAR imagery allows one to infer the bottom topography. The topography of the ocean floor can be mapped using the very precise ERS Altimeter, because the sea bottom relief is reflected on the surface by small variations of the sea surface height.
The ocean waves and their direction of displacement can be derived from the ERS SAR sensor operated in "Wave Mode". This provides input for wave forecasting and for marine climatology.
At high latitudes, SAR data is very useful for regional ice monitoring. Information such as ice type and ice concentration can be derived and open leads detected. This is essential for navigation in ice-infested waters.
On the land:
The ability of SAR to penetrate cloud cover makes it particularly valuable in frequently cloudy areas such as the tropics. Image data serve to map and monitor the use of the land, and are of increasing importance in forestry and agriculture.
Some geological or geomorphological features are enhanced in radar images thanks to the oblique viewing of the sensor and to its ability to penetrate (to a certain extent) the vegetation cover.
SAR data can be used to georeference other satellite imagery to high precision, and to update thematic maps more frequently and cost-effectively, due to its availability regardless of weather conditions.
In the aftermath of a flood, the ability of SAR to penetrate clouds is extremely useful. Here SAR data can help to optimize response initiatives and to assess damages.
Interferometric SAR (InSAR) can be used, under suitable conditions, to derive elevation models or to detect small surface movements, in the order of a few centimeters, caused by earthquakes, landslides or glacier advancement. This interferometric technique has strengthened as a result of the first ERS-1/ERS-2 Tandem phase, which lasted for about 9 months (until May 1996).
Remote Sensing Applications in the Earth Environment
Climate monitoring
Climate monitoring concerns the monitoring of the atmosphere and of other components of the earth system as well as the monitoring of global climate indicators (e.g. global mean earth surface temperature and precipitation). Satellite measurements appear to satisfy the need for global measurements.The earth climate shows great variability over different time scales spanning from decades to thousands of years and more. Past climatic conditions are studied by analysing ice cores, sea/lake sediments, shorelines movements, tree pollen, etc. Numerical experiments are also run in which a Global Circulation Model is used to explore the possible climatic changes related to, for example, the Earth axis oscillations. Knowledge of past climate can help in predicting the future. Abrupt changes may serve in the identification of thresholds values that can trigger a non-linear behaviour of the earth system (and hence may cause high variations). The overlapping of climate variability on different time scales is the very challenge in predicting climatic changes.A fundamental role in the determination of the earth climate is played by the solar radiation reaching the earth affecting the ground surface energy balance. The radiation spectrum at the earth is strongly influenced by atmospheric constituents: not only the amount of radiation but also its spectral distribution is crucial.
Coastal zone monitoring
Detection of oil spills
One of the most significant environmental concerns worldwide stems from oil pollution. During the last thirty years, pollution of the world's oceans, particularly in coastal areas, has become a matter of increasing international concern. In spite of rigorous controls, deterioration of water quality, especially in waters subject to heavy shipping, continues at a high rate. Due to the relative volumes of discharges, illegal emissions from ships represent a greater long-term source of harm to the environment than infrequent large scale accidents. Monitoring illegal discharges is thus an important component in ensuring compliance with marine protection legislation and the general protection of coastal environments. Traditionally, this service uses airborne patrols which are expensive and provide often only patchy coverage. Fast delivery SAR products are proving to be of great value in the optimisation of air-borne surveillance resources, due to the large area they can image at any one time. Size, location and dispersement of the oil spill can be conveniently determined using this type of imagery.
The "Sea Empress", a 147,000 ton supertanker, ran aground on rocks in the south of Wales, on the evening of February 15th, 1996. Seven days later, RADARSAT captured this image, clearly delineating the remaining oil slick. Size, location and disperse-ment of the oil spill can be conveniently determined using this type of imagery. The spill appears on the image in black tones.
Ship detection in coastal regions----Knowledge of the whereabouts and activities of ships in coastal regions is useful to a range of government and law enforcement agencies, such as those concerned with enforcing legislation regarding fishing activities in Exclusive Economic Zones, and environmental protection agencies to support pollutioncontrol. The information is also of use to the coastguard for use both in search and rescue operations and in law enforcement activities, to supplement land-based coastal surveillance radar which has a maximum range of under 100 km. It has long been recognised that satellite-based radar has the ability to detect and monitor vessel traffic. Due to the nature of the radar, monitoring can take place through cloud cover and at night thus proving an advantage over optical data. As well as detection of vessels it is possible to derive various characteristics of each vessel such as location, speed, heading, and broad class of vessel.
Land use, forestry and agriculture
In the original mission objectives, observing the land surface was viewed as an experimental application for ERS-1 data. However, the ability to monitor crop development and forestry changes independent of weather conditions, offers a major potential application area for ERS data.
An important technique which has been developed for terrestrial applications is multitemporal SAR analysis. Three input SAR datasets, acquired at different times, are assigned the colours red, green or blue. Changes between acquisitions can then be detected by observing the colours that appear in the image which reflect the change in the state of land cover. Crops planted at varying times and developing at varying rates can be identified, increasing the accuracy with which crop areas can be mapped and acreage estimated. Multitemporal analysis is also being applied to monitor logging in forested areas.
USES
Satellite photography can be used to produce composite images of an entire hemisphere...
...or to map a small area of the Earth, such as this photo of the countryside of Haskell County, Kansas, United States.
Satellite images have many applications in agriculture, geology, forestry, biodiversity conservation, regional planning, education, intelligence and warfare. Images can be in visible colours and in other spectra. There are also elevation maps, usually made by radar imaging. Interpretation and analysis of satellite imagery is conducted using software packages like ERDAS Imagine or ENVI. Some of the first image enhancement of satellite photos was conducted by the U.S. Government and its contractors. For example ESL Incorporated developed some of the earliest two dimensional Fourier transforms applied to digital image processing to address NASA photos as well as national security applications. Satellite imagery is also used in seismology and oceanography in deducing changes to land formation, water depth and sea bed, by colour caused by earthquakes, volcanoes, and tsunamis.
DISADVANTAGE
Because the total area of the land on Earth is so large and because resolution is relatively high, satellite databases are huge and image processing (creating useful images from the raw data) is time-consuming. Depending on the sensor used, weather conditions can affect image quality: for example, it is difficult to obtain images for areas of frequent cloud cover such as mountain-tops.
Commercial satellite companies do not place their imagery into the public domain and do not sell their imagery; instead, one must be licensed to use their imagery. Thus, the ability to legally make derivative products from commercial satellite imagery is minimized.
Privacy concerns have been brought up by some who wish not to have their property shown from above. Google Maps responds to such concerns in their FAQ with the following statement: "We understand your privacy concerns... The images that Google Maps displays are no different from what can be seen by anyone who flies over or drives by a specific geographic location."
Conclusions
It has been shown that there is a wide range of applications for satellite imaging radar products. Furthermore, ongoing research and development is continually expanding the current range of applications.one of the most important characteristics of imaging radars is their ability to penetrate cloud cover and to acquire data either by day or by night. It is this all-weather capability that has contributed significantly to the many commercial applications of satellite imaging radar.
