Microwave sensors (present and future)

Transcription

1 Proc. Indian Acad. Sci. (Engg. sea.), Vol. 6, Pt. 2, June 1983, pp Printed in India. Microwave sensors (present and future) 1. Introduction O P N CALLA Communications Area, Space Applications Centre, Indian Space Research Organisation, Ahmedabad , India Abstract. Microwave sensors with their unique capabilities offer certain specific advantages for applications like geological survey for petroleum and mineral prospecting, crop and vegetation monitoring, soil moisture detection, water resources management etc. They are also capable of measuring ocean surface wind speed and direction, wave heights, wave spectra etc and to map the ocean wave structure. This paper presents the basic physics and description of microwave remote sensing, both passive and active, sensors and reviews the present status and future possibilities in sensor technology and applications. Keywords. Remote sensing; microwave; radiometer; scanning radiometer; side looking radar; synthetic aperture radar; scatterometer; altimeter; soil moisture; crop classification. Remote sensing of earth resources using spaceborne sensors is becoming a useful tool of the application scientists to provide inputs to the natural resources monitoring and management systems. In the recent past, apart from optical sensors, sensors operating in the microwave and millimeter wave frequency bands also have been used in a limited scale for a variety of applications. These frequencies have certain unique capabilities such as propagating through clouds, penetrating soils and vegetation cover etc. Also they do not depend on sun's illumination and therefore can operate during day and night. With the advancement in signal processing technology, resolution comparable to that of optical sensors has become possible even from spaceborne platforms. Microwave sensors can be broadly classified as passive and active sensors. 1.1 Passive sensors All natural materials emit electromagnetic radiation which is a complex function of the physical properties of the emitting surface. Passive sensors are radiometers, which detect this radiated energy in the microwave spectrum. Useful information regarding physical properties as well as activities of the targets can be derived from the energy received. 1.2 Physical background The electromagnetic emission from a black body at a given temperature T~ governed by the Planck's law: is 2hf 3 B~ (2, T) = --~ [exp(hf/kt) - 1] - ~, (1) 109

2 110 0 P N Calla where B b is the spectral radiance (Wm- 2 Sr- ~ Hz- ~), Tthe physical temperature (~ C the free-space velocity of light (m/s), fthe frequency (Hz), K the Boltzmann constant (1-38 x J/~ and h the Planck's constant (6"63 x J.s). Figure I shows the variation of spectral emittance at different temperatures. It can be seen that radiation peaks at shorter wavelengths as the temperature is raised. One can show that for wavelengths longer than 1 mm and at all temperatures occurring in terrestrial environments (fits, 2 x 10x~ the black body radiation can be approximated by the Rayleigh-Jeans approximation: Bb(2, T) = 2KT/2 2. (2) For real media, the ability to absorb or emit is related by Kirchhoff's radiation law and therefore for non-perfect radiators, (2) can be suitably modified as B(2, T) = ~(2)(2KT/22), (3) where the emissivity e (2) is the ratio between the emission from the object and the black body maintained at the same thermodynamic temperature. The emissivity depends on a number of parameters such as temperature, polarisation, frequency, angle of incidence and physical properties of the surface. In the case of land surface, emissivity depends strongly on the surface features as well as moisture content in the soil. Similarly for ocean surface, it depends on the temperature, sea state, salinity etc. Figure 2 shows typical emissivity variation due to different soil moisture content. The emission from objects in natural environments is random and therefore the intervening atmosphere modifies the signal strength due to attenuation as well as selfemission. The atmospheric attenuation is caused due to transition between molecular states of gaseous constituents as well as absorption and scattering due to the water droplet, hail and precipitation. "i N 108 E 'E m m m t- 'r" "~ -24 lo 10 million "K/~ "K~ / looy// /I frequency~ Hz 'Tk_ tr) 1040 a E 7 E 1030 O 102o.~-- v e~ 1010,n 1 5 lo-lo ~ Figure 1. Variation of spectral emittance at different temperatures.

3 Microwave sensors 111 A v U l.0.'2_ "g % moisture vertical polarization I I I I angle of incidence (deg) nr 1.0 v u "'~'~~2~ moisture "~ 0.6 tn gi ~horizontal - "~..=3o "/. "\~ 2 polarization I I I I angle of incidence (deg) Figure 2. Brightness temperature variations with soil moisture t (lntennct radiometer output reference termir~tion Figure 3. Block diagram of radiometric receiver. LO: local oscillator; synch det: synchronous detector; amp: amplifier. 1.3 System description Radiometers, which are highly sensitive receivers are of many types. The most extensively used is the Dickr radiometer which contains an antenna (which should have (Enoo. Sci.)--2

4 112 0 P N Calla low loss and low sidelobes), mixer, IF amplifier, square law detector and DC amplifier. A Dicke switch is used to switch the receiver to the antenna and the reference load. The temperature sensitivity of the radiometer depends on the receiver noise figure, the system bandwidth and the integration time. Figure 3 is a simplified block diagram of a typical radiometer. To cover a wider area, scanning antennas are used. Scanning is either by mechanical movement of the antenna or by phased array antenna. 1.4 Present status and future trends Passive microwave measurements from earth orbits provide global data on a wide range of geophysical and meteorological phenomena. Though the technique was developed during the thirties and forties, it was effectively used in remote sensing only in 1967 when electronically scanned microwave radiometer was flown on-board NASA CV 990 aircraft. Later the same equipment was orbitted onboard Nimbus-5 satellite. The first spacecraft radiometer operating at 1.35cm and 1.9 cm wavelengths was launched on-board Mariner-ll, Venus fly by mission. However, the first multichannel radiometer for remote sensing of geophysical parameters was orbitted on-board Soviet Cosmos-243 and Cosmos-384 in 1968 and 1970 respectively. The Cosmos experiments included measurements at 8.5, 3.4, 1.31 and 0.8cm wavelenghts for observing polar ice, sea surface temperature and atmospheric water vapour and liquid water over oceans. Subsequently, a large number of radiometers have been flown for a variety of applications. A two-frequency radiometer operating at 19 and 22 GHz and a three-frequency radiometer operating at 19, 22 and 31 GHz were developed at the Space Applications Centre, Ahmedabad and were flown successfully on-board Bhaskara-I and Bhaskara- II, experimental earth observation satellites, launched by the Indian Space Research Organisation during 1979 and 1981 respectively. The system specifications are given in table 1. The data obtained have been satisfactorily used to study oceans and land- Table!. Samir payload specifications Bhaskara-I Bhaskara-ll Frequency (GHz) RF Bandwidth (MHz) System noise figure (max.) (db) Predetection bandwidth (MHz) Integration time (ms) Temperature sensitivity (~ (max.) 1 1 Antenna beamwidth (deg) 16 ~ for 14 ~ for all R1 & R2 channels 26 ~ for R3 Input brightness temperature range (~

5 Microwave sensors 113 Figure 4. Radiometer developed for Bhaskara spacecraft related phenomena. Figure 4 is a photograph of the radiometer developed for Bhaskara. The main thrust at present is to improve spatial resolution, specially for land-related applications. Extensive studies for this purpose have been conducted by the National Aeronautics and Space Administration (NASA), by having very large antenna to be assembled in space. The antenna diameter envisaged is from 45 to 725 m to provide a resolution capability around 1 to 2 kin. 2. Active sensors As seen earlier the passive sensors use natural emission from the targets to identify and study their physical characteristics. In the case of active sensors, the sensor illuminates the targets with its own energy. The energy thus scattered in the direction of the sensor is processed to study their physical properties and activities as well as to generate imageries of the targets as a function of their physical and electrical properties like size, shape, conductivity and dielectric constant as well as frequency, polarisation and the illumination angle of the sensor. The power received by the sensor for a given transmitted power and range is expressed as P, oc ~ro A, (4) where ao is the scattering coefficient and A is the physical area of the target. As each

6 114 0 P N Calla natural object has a different scattering coefficient depending on their physical and electrical properties, One can separately identify them and study their status, by measuring the scattered power. Another important property of the microwaves is to penetrate soil and vegetation cover. At very high microwave frequencies (about 10 GHz) the current flows mainly on the surface. But at lower frequencies, the current flow takes place in the subsurface also. This is determined by the 'skin depth' which is a function of frequency, conductivity, permittivity etc. This property makes it possible to study the soil moisture in the subsurface and see through the vegetation and soil cover for geological and water resources application. There are different active sensors for different applications. The most popular among them are side looking radar, scatterometer and altimeter. 2.1 Side-lookino radar Side-looking radar (SLR) is an imaging radar used for reconnaissance applications. However, it has tremendous potential for remote sensing of natural resources, monitoring of flood and other natural calamities, ocean pollution, oceanographic studies etc. It consists of a transmitter to transmit high power pulsed microwave energy at a given frequency via a fan beam antenna which illuminates the targets/terrain. The energy scattered back by the targets/terrain is received by the same antenna and coupled to the receiver, which processes the signal and the image is generated. High resolution is obtained through narrow transmitted pulse in the across track direction and through the narrow antenna beam in the along-track direction as defined below across track resolution = C~/2 cos 0, (5) along track resolution = Rfl, (6) where z is the pulse width, 0 is the illumination angle or the incident angle, R is the range from radar to target and fl is the beam width of the antenna. One can see that there is a limit to the along track resolution achievable specially at very high ranges. In order to overcome this limit, the synthetic aperture concept is used. In this case, the amplitude and phase information of the targets as they traverse the beam in the along-track direction is received and processed to achieve very fine along-track resolution. By this technique the along track resolution achievable is L/2 where L is the length of the antenna and the resolution is independent of the range. Figure 5 gives the geometry of SLR and figure 6 the simplified block diagram. 2.2 Scatterometer The scatterometer is another form of radar used to measure the scattering coefficient of the targets. It consists of a transmitter transmitting pulsed or FM-CW signal. The transmit and receive antenna can be a fan beam or a pencil beam. In the receiver, the received signal is processed to compute the scattering coefficient of the targets. Here the

7 Microwave sensors 115 a L t O " _ antenna beamwidth in along track direction e- incident or illumination angle R -- Range ij~ swath =- I I I I I IIIII 1 11, I, 2s to= Figare 5. Basic geometry of side looking radar )ac =across track resolution C~ 2 cos e = a along track L resolution = Rp transmitter ~ ontennct SCM I Jscan converter c~176 1 memory(scm) I Figure 6. Simplified schematic diagram of SLAR emphasis is on absolute calibration for better confidence level on the measured value unlike in SLR where relative calibration is sufficient for most applications. Scatterometers are used as ground-based instruments to measure scattering coefficients of various targets like crops, vegetation, soils et~., under controlled

8 116 0 P N Calla environment. By carrying out this measurement for different frequency, polarisation and look angle, one can arrive at appropriate sensor parameters. Also this data can be used for analysing the radar data using the computer. Scatterometers are also used from aircraft and spacecraft to measure wind speed and direction over the ocean, as there is a good correlation between the scattering coefficient and wind speed and direction. 2.3 Altimeter Altimeter is also a radar but used for measuring altitude. The delay between the transmitted and received signal is processed to get the distance information. This principle is used for measuring wave heights, geoids etc. 2.4 Present status and future trends 2.4a Side-looking radars Side-looking airborne radar (SLAR) operating in real aperture as well as synthetic aperture radar (SAR) mode has been used extensively to study its remote sensing applications. SEASAT-A, a satellite launched by NASA in 1978 carried a SAR operating at L-band. It produced imagery/data with 25 m resolution and 100 km swath. From aircraft, resolutions of the order of 3 m and better have been achieved. An airborne SEa operating at 9"65 GHz has been developed at the Space Applications Centre (SAC), Ahmedabad. This produces imageries with 25 m resolution from an altitude of 2 km. A similar system has also been developed at NRSA, Hyderabad. Figure 7 is a photograph of the system mounted inside a DC-3 aircraft. 2.4b Scatterometer Scatterometers for ground-based measurements have been developed by different agencies. Notable among them is the 2-18 GHz FM-CW scatterometer developed at the University of Kansas, USA. This can measure scattering coefficient at various combinations of frequency, polarisation and look angle automatically, compute the scattering coefficient and print it out in real time. A similar unit is under development at SAC, Ahmedabad. The 8-18 GHz unit has already been developed and integrated with a mobile platform for field measurements. For oceanographic applications like wind speed and direction measurements, Skylab and SEASAa" carried scatterometers which provided data to measure wind speed over a range of 5 m/s to 30 m/s with an accuracy of _ 2 m/s. 2.4c Altimeter Altimeters also have been flown on Skylab, Geos-3 and SEASAT-A. The SEAS^T altimeter measured wave heights with an accuracy _+ 10 cm accuracy. 2.5 Future trends In order to properly interpret and maximise information extraction, it is important that data should be gathered at more than one frequency and polarisation. Often the crosspolarised components give better insight. Obviously the effort should be to make sensors with multifrequency and cross-polarisation capabilities. One of the major constraints here is the antenna. It is a challenging job to have low sidelobes and cross-

9 Microwave sensors 117 Figure 7. SLAR system mounted inside DC-3 aircraft polarisation components and at the same time operate at more than one frequency and polarisation. Another important area for further work is to achieve finer image resolution or signal resolution. As far as spatial resolution is concerned, very fine resolution capability has been established using synthetic aperture technique. But the signal resolution required is of the order of and _ 1 db for applications like soil moisture, crop yield monitoring etc. and this is yet to be achieved. For signal processing, the future trend is to process the signal in real time using digital and analog processing techniques. On-board signal processing and direct data transmission to the user are also major goals for the future. Presently SAR systems operating at L-band are being flown on-board Shuttles, the reusable launchers of NASA, USA. The Spacelab to be launched will also carry an X-band SAP,. The European Space Agency (ESA) and the Japanese Space Agency (JSA) are likely to launch spaceborne SARS during the second half of this decade. 3. Applications of microwave sensors This section gives briefly the various applications of microwave sensors for remote sensing and their present status and future possibilities.

10 118 0 P N Calla Table 2. Potential applications 1. Agriculture --Crop classification --Crop yield monitoring --Soil moisture measurement --Cultivated and forest area mapping 2. C~mlogy --Petreleum and mineral prospecting --Large scale mapping of nearly forested geological sites 3. Water resources --Mapping of water bodies ---Snow cover mapping --Drainage pattern study 4. Flood mapping 5. Land use survey 6. Coastal monitoring 7. Oceanography ---Surface wind speed and direction --Wave height and wave spectrum 8. Environmental studies The potential applications of microwave sensors based on certain experimental studies and on their unique capabihties are given in table 2. Among these, geological applications, land use applications, oceanographic applications like wind speed, direction and wave height measurements etc have been fairly well established. However for applications like soil moisture, crop classification, crop yield prediction etc., though the potentiality has been established, there is a long way to go before one can reach the operational situation. As the radar illuminates the targets in oblique angles, terrain features like slopes, lineaments etc come out very well on the radar imagery, due to the shadow effect. Also the capability to penetrate the surface soil and vegetation helps to map the geological sites covered with vegetation and soil. These reasons as well as the capability to produce photography-like imagery have helped in using the SLR data extensively for geological and land use applications. Mapping of water bodies, drainage patterns etc are also effectively done by microwave sensors due to the vast difference in their scattering coefficient. Experiments in North America and Europe have established their potential for crop classification and delineations of cultivated and forest areas. The unique application of microwave sensor is the possible detection of soil moisture at surface and subsurface. Experiments at the University of Kansas have estabhshed a good correlation between soil moisture and the scattering coefficient. The optimum frequency has been found to be 4 to 5 GHz. As the microwave sensor is an all-weather sensor, it is uniquely suited to provide timely information for agricultural applications. Hence, the thrust in the near future will be to establish these applications in operational mode. For this purpose, extensive signature studies and modelling studies will be required.

11 4. Conclusions Microwave sensors 119 Due to their unique capabilities like all-weather and day and night operation, detection of moisture in soil, snow and vegetation and penetration of soil, they are suitable for agricultural and geological applications. The technology elements of these sensors are fairly well understood, though there are some more challenges to be met. With the advancement in signal processing techniques and further insight into the mechanism of electromagnetic interaction, it will be possible in future to realise some of these potential applications in operational situations. The author acknowledges the assistance received from Shri N. S. Pillai, Shri S. S. Rana, Shri G. Raju and Shri T. R. Ramachandran.