Microwave remote feeling at wavelengths runing from 1 centimeter to 1 m has gained a batch of importance over the plast decennary for a broad scope of scientific applications with the handiness of active radio detection and ranging imaging systems. Its possible in spacial applications usage has been scientifically established in assorted sectors like forestry, agribusiness, land usage and land screen, geology and hydrology.
A assortment of applications have been carried out universe over utilizing microwave informations like favoritism of harvest types, harvest status monitoring, dirt wet retrieval, word picture of forest gaps, appraisal of forest above land biomass, forest function ; forest construction and fire cicatrix function, geological function, supervising wetlands and snow screen, sea ice designation, coastal windfield measuring, wave incline measuring, ship sensing, shoreline sensing, substrate function, slick sensing and general flora function ( Kasischke et al. , 1997 ) .There is an emerging involvement on microwave remote detection is, as microwave detectors it can image a surface with really all right declaration of a few metres to coarse declaration of a few kilometres. They provides imagination to a given declaration independently of height, limited merely by the sender power available. Cardinal parametric quantities like polarisation and look angle can be varied to optimise the system for a specific application. SAR imagination is independent of solar light as the system provides its ain beginning of light. It can run independently of conditions conditions if sufficiently long wavelengths are chosen. It operates in a set of electromagnetic spectrum different from the sets used by seeable and infrared ( IR ) imaginations.
Microwave applications in Forestry
Applications of microwave remote feeling in forestry hour angle ve besides been reported during the recent yesteryear. Recent reappraisals on the application of radio detection and ranging in forestry show that SAR systems have a good capableness in know aparting assorted types of ( tropical ) forest screen utilizing multi-temporal and multi-frequency SAR information ( Vander Sanden, 1997 ; Varekamp, 2001 ; Quinones, 2002 ; Sgrenzaroli, 2004 ) . These surveies showed that the biomass dependance of radio detection and ranging backscatter varies as a map of radio detection and ranging wavelength, polarisation and incidence angle. Besides recent surveies have demonstrated that man-made aperture radio detection and ranging ( SAR ) can be used to gauge above-ground standing biomass. To day of the month, these surveies have relied on extended ground-truth measurings to build relationships between biomass and SAR backscatter ( Steininger, 1996 ; Rignot et al.
, 1997 ) .Many surveies demonstrated the usage of Man-made Aperture Radar ( SAR ) remote feeling to recover biophysical features from forest marks ( Richards, 1990 ) . Although radio detection and ranging backscatter from wood is influenced by their structural belongingss ( Imhoff, 1995 ) , earlier surveies derived utile relationships between backscattering coefficients and the above-ground biomass ( Baker et. , 1994 ; Le Toan et al. , 1992 ; Dobson et el. , 1992 ; Imhoff ; 1995 ) . These relationships may supply a method of supervising forest ecosystems which play such a critical function in C storage and NPP.
Microwave remote detection has the advantage of all conditions capableness coverage get the better ofing the relentless job of cloud screen in satellite images like in optical informations. Optical remote detection is being used really successfully in assorted applications related to earth resources surveies and monitoring of the environment. However, optical remote detection is non suited for all atmospheric conditions. It can non perforate through clouds and haze.
In many countries of the universe, the frequent cloud conditions frequently restrain the acquisition of high-quality remotely sensed informations by optical detectors. Therefore, radio detection and ranging information has become the lone executable manner of geting remotely sensed informations within a given clip model because the radio detection and ranging systems can roll up Earth characteristic informations irrespective of conditions or light conditions. Due to this alone characteristic of radio detection and ranging informations compared with optical detector informations, the radio detection and ranging informations have been used extensively in many Fieldss, including forest-cover designation and function, favoritism of forest compartments and forest types, appraisal of forest base parametric quantities and monitoring of woods.
In countries where flora screen is heavy, it visually covers the implicit in formation and it is really hard to observe structural restricting the usage of optical detectors. Radar nevertheless, is sensitive plenty to topographic fluctuation that it is able to spot the structural look reflected in the tree top canopy, and hence the construction may be clearly defined on the radio detection and ranging imagination.Based on this background, the current thesis work has been carried out to research the potency of microwave informations in turn toing nucleus countries of tropical forestry viz. , flora categorization, a bove land biomass appraisal etc.
, and to supply the users/researchers a meaningful information base of SAR applications in tropical forestry, specifically over the India part.
- Which SAR wavelength/frequency set is appropriate for flora categorization in tropical woods?
- To what extent above land biomass can be measured in tropical woods?
- Which frequence set and polarisation are suited for above land biomass appraisal?
- Is there any sweetening in flora categorization with polarimetric / interferometric informations than stand alone amplitude informations?
Based on this background, the old surveies and before mentioned Rresearch inquiries, we understand that the backscatter increases with the addition in above land biomass and depends on wavelength sets, polarisations used and on the survey country, topographic fluctuations and species composing. So, the present survey efforts to deduce the application potency of airborne and infinite borne SAR information in the quantification of the forest resources in tropical parts like India, both as a complementary and auxiliary function to optical datasets. Different techniques such as Regression analysis, multi-sensor merger, texture steps and interferometric coherency qualify different biomass scopes of the trial sites and categorization of major land screen classes. This survey would ease range for future research in tropical parts to research the potencies of SAR informations in land screen categorization and above land biomass appraisal utilizing the polarimetric and interferometric techniques.
Based on this background, the present survey purposes at the following aims:
- Vegetation type categorization utilizing polarimetric and interferometric SAR information.
- Forest above-ground biomass appraisal utilizing multi-frequency SAR informations and land inventoried informations.
Vegetation categorization is necessary to understand the diverseness of species in a given country which gives above land biomass with mensural parametric quantities.
Hence, flora categorization enhances the appraisal of the above land biomass.Forest biomass is a cardinal parametric quantity in understanding the C rhythm and finding rates of C storage, both of which are big uncertainnesss for forest ecosystems. Accurate cognition of biophysical parametric quantities of the ecosystems is indispensable to develop an apprehension of the ecosystem and their interactions, to supply input theoretical accounts of ecosystem and planetary procedures, to prove these theoretical accounts and to supervise alterations in ecosystem kineticss and processes over clip. Therefore, it is a utile step for measuring alterations in wood construction, comparing structural and functional properties of forest ecosystems across a broad scope of environmental conditions.
Knowing the spacial distribution of forest biomass is of import as the cognition of biomass is required for ciphering the beginnings and sinks of C that consequence from change overing a wood to cleared land and frailty versa, to cognize the spacial distribution of biomass which enables measuring of alteration through clip.Field sampling is the most followed conventional method for flora type categorization. The designation of different species in field outputs good consequences in the appraisal of the above land biomass. It is really clip consuming, expensive and really complicated.
With the usage of multiple detectors, varied informations aggregation and reading techniques, distant detection is a various tool that can supply informations about the surface of the Earth to accommodate any demand ( Reene et al, 2001 ) . Distant feeling attack for flora categorization is cost effectual and besides clip effectual. Though the designation of the tree species is possible merely from the aerial imagination, major forest types can be identified from the airborne and the spaceborne remote feeling informations. Ocular image reading provides a executable agencies of flora categorization in woods.
The image features of form, size, form, shadow, tone and texture are used by translators in tree species designation. Phenological correlativities are utile in tree species designation. Changes in the visual aspect of trees in different seasons of the twelvemonth some times enable favoritism of species that are identical on individual day of the months. The usage of multi-temporal distant feeling informations enables the function of the different wood types.SAR has shown its potency for sorting and supervising geophysical parametric quantities both locally and globally. Excellent plants were carried out on the categorization utilizing several attacks such as polarimetric informations decomposition ( Lee et al. , 1998 ) , knowledge based attacks sing the theoretical backscatter mold and experimental observations ( Ramson and Sun, 1994 ) ; Backscatter model-related inversion attacks ( Kurvonen et al. , 1999 ) , nervous webs and informations merger attacks ( Chen et al.
, 1996 ) . Dong et Al. ( 2001 ) have shown that the categorization truth of 95 % for the flora categories could be achieved through the cleavage and categorization of the SAR informations utilizing Gaussian Markov Random Field Model ( GMRF ) .Many methods have been employed for categorization of polarimetric SAR information, based on the maximal likeliness ( ML ) ( Lee et al. 1994 ) , unreal nervous web ( NN ) ( Chen et al. 1996, Ito and Omatu, 1998 ) , support vector machines ( SVMs ) ( Fukuda et al.
2002 ) , fuzzed method ( Chen et al. 2003, Du and Lee 1996 ) , or other attacks ( Kong et al. 1988, Lee and Hoppel 1992, van Zyl and Burnette 1992, Cloude and Pottier 1997, Lee et al. 1999, Alberqa 2004 ) Among these methods, the ML classifier ( Lee et al.
1994 ) can be employed for obtaining accurate categorization consequences, but it is based on the premise of the complex Wishart distribution of the covariance matrix.Measuring the entire aboveground biomass of woods ( biomass denseness when expressed as dry weight per unit country at a peculiar clip ) is a utile manner of quantifying the sum of resource available for all traditional utilizations. It either gives the measure of entire biomass straight or the measure by each constituent ( e.g.
, leaves, subdivisions, and bole ) because their biomass tends to change consistently with the entire biomass. However, biomass of each constituent varies with entire biomass by forest type, such as natural or deep-rooted woods and closed or unfastened woods. For illustration, leaves contribute about 3-5 % and marketable bole is about 60 % of the entire aboveground biomass of closed woods.Many research workers have developed assorted methods based on field stock list and remote feeling attacks for the appraisal of above land biomass ( Kira and Ogawa, 1971 ) . Traditionally, field-measured attack is considered as the most accurate beginning for above-ground biomass appraisal. It has been converted to volume, or biomass, utilizing allometric equations that are based on standard field measurings ( tree tallness and diameter at chest tallness ) .Different attacks, based on field measuring ( Brown et al. 1989, Brown and Iverson 1992, Schroeder et al.
. 1997, Houghton et al. , 2001, Brown, 2002 ) ; remote detection ( Tiwari 1994, Roy and Ravan 1996, Tomppo et al. , 2002, Foody et al. , 2003, Santos et al.
, 2003, Zheng et al. , 2004, Lu, 2005 ) ; and GIS ( Brown and Gaston 1995 ) have been applied for AGB appraisal. Traditional techniques based on field measuring are the most accurate ways for roll uping biomass informations.
A sufficient figure of field measurings is a requirement for developing AGB appraisal theoretical accounts and for measuring the AGB appraisal consequences. However, these attacks are frequently clip devouring, labour intensifier, and hard to implement, particularly in distant countries and are by and large limited to 10-year intervals. Besides, they can non supply the spacial distribution of biomass in big countries.For the above grounds, the positions of utilizing distant feeling techniques to gauge forest biomass have gained involvement. Remote feeling informations available at different graduated tables, from local to planetary, and from assorted beginnings, optical to micro-cook are expected to supply information that could be related indirectly, and in different manners, to biomass information. The possibility that aboveground forest biomass might be determined from infinite is a promising alternate to ground-based methods ( Hese et al.
, 2005 ) .The advantages of remotely sensed informations, such as in repetivity of informations aggregation, synoptic position, digital format that allows fast processing of big measures of informations, and the high correlativities between spectral sets and flora parametric quantities, make it the primary beginning for big country AGB appraisal, particularly in countries of hard entree. Therefore, remote sensing-based AGB appraisal has progressively attracted scientific involvement.
In general, AGB can be estimated utilizing remotely sensed informations with different attacks, such as multiple arrested development analysis, K nearest-neighbour, and nervous web ( Roy and Ravan 1996, Nelson et Al. 2000a, Steininger 2000, Foody et al. 2003, Zheng et Al.
2004 ) , and indirectly estimated from canopy parametric quantities, such as crown diameter, which are first derived from remotely sensed informations utilizing multiple arrested development analysis or different canopy coefficient of reflection theoretical accounts ( Wu and Strahler 1994, Woodcock et Al. 1997, Phua and Saito 2003, Popescu et Al. 2003 ) .Spectral signatures or flora indices are frequently used for AGB appraisal in optical remote feeling.
Many flora indices have been developed and applied to biophysical parametric quantity surveies ( Anderson and Hanson 1992, Anderson et Al. 1993, Eastwood et Al. 1997, Lu et Al. 2004, Mutanga and Skidmore 2004 ) . Vegetation indices have been recommended to take variableness caused by canopy geometry, dirt background, sun position angles, and atmospheric conditions when mensurating biophysical belongingss ( Elvidge and Chen 1995, Blackburn and Steele 1999 ) .Radar remote detection has possible to supply information on above land biomass. The information content of SAR informations in footings of the retrieval of biomass parametric quantities will be assessed based on an apprehension of the underlying sprinkling mechanisms, which in bend are derived from observations and patterning consequences.
For this intent, an analysis of informations acquired by multiple frequence, incidence and polarization systems and by interferometric systems is carried out. It has been proved that the sensitiveness to biomass parametric quantities differ strongly at different frequences, polarizations and incidence angles.In general, long wavelength SAR backscatter ( P and L set ) is more sensitive to forest biomass than shorter wavelength C-band backscatter and the relationships saturate at certain biomass degrees ( Imhoff 1995b ) . The strength of the relationships and the impregnation degrees are dependent on the type of forest being analysed ( Ferrazoli et al. 1997 ) . The impregnation degrees for the appraisal of above land biomass depend on the wavelengths ( i.e.
different sets, such as C, L, P ) , polarisation ( such as HV and VV ) , and the features of flora base construction and land conditions. C-band can mensurate forestry biomass up to app. 50 tons/ha, L-band can mensurate up to 100 tons/ha and P-band can mensurate up to 200 tons/ha ( Floyd et al. , 1998 ) . The combination of multiple channels and polarisations provides greater advantage for gauging entire biomass ( Harry Stern, 1998 ) .
Relevance OF THE STUDY:
The present survey is the portion of Radar Imaging orbiter – Joint Experiment Programme ( RISAT-JEP ) for forestry applications undertaken by Forestry and Ecology Division of National Remote Sensing Centre ( NRSC ) , as a pilot run with specific aims of above land biomass appraisal and flora type categorization utilizing airborne DLR ( German Aerospace Center ) transporting ESAR ( Experimental Synthetic Aperture Radar ) information for Rajpipla ( Gujarat ) survey site and infinite borne ENVISAT ( Environmental Satellite ) transporting Advanced Synthetic Aperture Radar ( ASAR ) information for three trial sites viz. , Rajpipla ( Gujarat ) , Dandeli ( Karnataka ) and Bilaspur ( Chattisgarh ) , India.
Scope OF THE STUDY:
The specific aims of the present survey are above land biomass appraisal and flora type categorization utilizing airborne DLR ( German Aerospace Center ) transporting ESAR ( Experimental Synthetic Aperture Radar ) information for Rajpipla ( Gujarat ) survey site and infinite borne ENVISAT ( Environmental Satellite ) transporting Advanced Synthetic Aperture Radar ( ASAR ) information ; ALOS ( Advanced Land Observing Satellite ) transporting Phased Array L-band Synthetic Aperture Radar ( PALSAR ) for three trial sites viz. , Rajpipla ( Gujarat ) , Dandeli ( Karnataka ) and Bilaspur ( Chattisgarh ) , India.Different techniques such as Regression analysis, multi-sensor merger, texture steps and interferometric coherency were used to qualify different biomass scopes of the trial sites and to sort the major land screen classes utilizing spaceborne C-band ENVISAT-ASAR informations and L-band ALOS- PALSAR informations.
Polarimetric signatures, polarimetric decompositions, multi-sensor merger techniques etc. were used for the categorization of different flora types in the Rajpipla survey country utilizing the airborne DLR-ESAR information.The survey has its uniqueness and additions importance in the application potency of SAR interferometry over tropical parts like India, both in footings of an alternate/substitute to optical informations sets due to prevailing cloud screen and to the deficiency of handiness of any earlier scientific work over the survey part. This survey is utile for the applications of to be launched Radar Imaging Satellite ( RISAT ) in 2010.The survey has richly demonstrated the application potency of airborne and infinite borne SAR information in the quantification of the forest resources in tropical parts like India, both as a complementary and auxiliary function to optical datasets. The survey would ease future research in tropical parts to research the potencies of SAR informations in land screen categorization and above land biomass appraisal utilizing the polarimetric and interferometric techniques.
During the last decennary, many possible applications of SAR in different frequence sets have been studied for forestry applications utilizing informations acquired by both airborne and space-borne systems. Assorted techniques like Polarimetry, Interferometry and Polarimetric-Interferometry enhanced the usage of SAR informations in forestry applications. The backscatter from flora is used to deduce information about amplitude informations for forest screen function and appraisal of above land biomass in renewing woods. Use of SAR polarimetric informations delineated flora categories within the wood and besides enhanced the capableness in gauging the above land biomass.
The usage of repetition base on balls interferometric informations enables to cipher the forest base tallness and besides used for the land screen categorization. The emerging Pol-InSAR technique is used to deduce the three dimensional forest constructions.Forest screen maps were prepared for the boreal, temperate and tropical woods utilizing SAR informations. Forest was separated from non-forest parts utilizing multi-temporal C-band ERS SAR informations on the trial sites of United Kingdom, Poland and Finland ( Quegan et al. , 2000 ) . The survey applied a threshold value to divide forest from other categories.
Tropical rain forest of Borneo was mapped from SIR-B informations of different incidence angles ( Ford and Casey, 1988 ) . Different flora covers along with wetlands and distinct countries were distinguished. Forest screen function was done with JERS-1 SAR informations on the coastal parts of Gabon ( Simard et al. , 2000 ) . The survey used determination tree method using both radio detection and ranging amplitude and texture information. Forest screen map was prepared for Southern Chittagong utilizing JERS-1 SAR informations ( Rahman and Sumantyo, 2007 ) and the survey separated forest, degraded forest, bush, coastal plantations, agribusiness, shrimp-farms, urban and H2O.Although radio detection and ranging backscatter from wood is influenced by their structural belongingss ( Imhoff, 1995a ) , many surveies have demonstrated utile relationships between backscattering coefficients and the areal denseness of above-ground biomass within peculiar types of wood ( Baker et. , 1994 ; Le Toan et al.
, 1992 ; Dobson et al. , 1992 ; Imhof et Al ; 1995b ) .Many airborne and spaceborne SAR systems have been used to transport out a big sum of experiments for look intoing the forest ecosystems.
The airborne systems, such as the NASA/JPL AIRSAR, DLR-ESAR, etc. , runing at P, L and C set, has been flown over many forest sites ( Zebker et al. , 1991 ; Le Toan et Al, 1992 ; Beaudoin et al. , 1994 ; Rignot et Al. ; 1994 ; Skriver et al.
, 1994 ; Ranson et al. , 1996 ) . The experiments of the Canadian CV-580, every bit good as the European airborne system, chiefly runing at C and X set besides have been carried out in North America and Europe ( Drieman et al. , 1989 ; Hoekman, 1990 ) . Spaceborne SAR is being used from regional to planetary monitoring in a periodic footing. The spaceborne systems, such as the Seasat SAR, SIR-B, SIR-C/X-SAR and ERS-1, ERS-2, ENVISAT-ASAR, RADARSAT etc. , were used for probes of boreal, temperature and sub-tropical forestry trial sites ( Ford et al. , 1988 ; Dobson et al.
, 1992 ; Ranson et al. , 1995 ; Stofan et al. , 1995 ; Rignotet al.
, 1995 ) . These experiments and surveies have shown that radio detection and ranging is sensitive to forest structural parametric quantities such as diameter at chest tallness ( dbh ) and tree mean height including above-ground biomass ( Dobson et al. , 1992 ; Pulliainen et al. , 1994 ; Skriver et al. , 1994 ; Ferrazzoli et al. , 1995 ; Ranson et al. , 1996 ) .
Earlier surveies has shown the potency of radio detection and ranging informations in gauging AGB ( Hussin et al. 1991, Ranson and Sun 1994, Dobson et Al. 1995, Rignot et Al. 1995, Saatchi and Moghaddam 1995, Foody et al. 1997, Harrell et Al. 1997, Ranson et Al. 1997, Luckman et al.
1997, 1998, Pairman et al. 1999, Imhoff et Al. 2000, Kuplich et al. 2000, Castel et Al.
2002, Sun et Al. 2002, Santos et al. 2003, Treuhaft et Al.
2004 ) . Kasischke et Al. ( 1997 ) reviewed radio detection and ranging informations for ecological applications, including AGB appraisal. Lucas et Al. ( 2004 ) and Kasischke et Al. ( 2004 ) reviewed SAR informations for AGB appraisal in tropical woods and temperate and boreal woods, severally. Different wavelength radio detection and ranging informations have their ain features in associating to forest base parametric quantities. Backscatter in P and L sets is extremely correlated with major wood parametric quantities, such as tree age, tree tallness, DBH, radical country, and AGB ( Leckie 1998 ) .
In peculiar, SAR L-band informations have proven to be valuable for AGB appraisal ( Sader 1987, Luckman et al. 1997, Kurvonen et Al. 1999, Sun et Al.
2002 ) . However, low or negligible correlativities were found between SAR C-Band backscatter and AGB ( Le Toan et Al. 1992 ) . Beaudoin et Al.
( 1994 ) found that the HH return was related to both bole and crown biomass, and the VV and HV returns were linked to crown biomass.Harrell et Al. ( 1997 ) evaluated four techniques for AGB appraisal in pine bases utilizing SIR C- and L-Band multi-polarization radio detection and ranging informations and found that the L-Band HH polarisation informations were the critical elements in AGB appraisal. Kuplich et Al. ( 2000 ) used L-band JERS-1/SAR informations for AGB appraisal of renewing woods and concluded that these informations had the possible to gauge AGB for immature, renewing woods. Sun et Al.
( 2002 ) found that multi-polarization L-Band SAR informations were utile for AGB appraisal of forest bases in cragged countries. Castel et Al. ( 2002 ) identified the important relationships between the backscatter coefficient of JERS- 1/SAR informations and the base biomass of a pine plantation. The survey observed the betterment in AGB appraisal consequences for immature bases, compared to appraisal for old bases. Santos et Al. ( 2002 ) used JERS-1 SAR informations to analyze the relationships between backscatter signals and biomass of wood and savanna formations.
This survey concluded that forest structural-physiognomic features and the radio detection and ranging ‘s volume sprinkling, dual bounciness dispersing are two of import factors impacting these relationships. The impregnation degrees of backscattering co-efficient with regard to AGB depend on the wavelengths ( i.e. different Bands, such as C, L, P ) , polarisation ( such as HV and VV ) , and the features of flora base construction and land conditions. Luckman et Al. ( 1997 ) found that the longer-wavelength ( L-Band ) SAR image was more suited to know apart different degrees of wood biomass up to a certain threshold, bespeaking that it is suited for gauging biomass of renewing woods in tropical parts.
Austin et Al. ( 2003 ) indicated that forest biomass appraisal utilizing radio detection and ranging informations may be executable when landscape features are taken into history.The radio detection and ranging backscattering coefficient is correlated with forest biomass and root volume ( Le Toan et Al. 1992, Israelsson et Al. 1994, Kasischke et Al.
1994, Dobson et Al. 1995 ) . The sensitiveness of Man-made Aperture Radar ( SAR ) information to forest root volume increases significantly as the radio detection and ranging wavelength additions ( Israelsson et al. 1997 ) .
The imaging procedure makes SAR suited for mapping parametric quantities related to forest biomass, like root volume ( Baker et al, 1999 ; Fransson et Al, 1999 ; Hyyppa et Al, 1997 ; Israelsson et al. , 1997 ; Kurvonen et Al, 1999 ; Pulliainen et Al, 1996 ) , entire turning stock ( Balzter et al, 2000 ; Schmullius et Al, 1997 ) , LAI ( Imhoff et al, 1997 ) , or above land net primary productiveness ( Bergen et al, 1998 ) .Le Toan et al. , ( 1992 ) used multi-polarisation L- and P-band airborne radio detection and ranging informations, and found that the dynamic scope of the radio detection and ranging backscatter corresponded extremely with forest growing phases and is maximal at P-band HV polarisation. The analysis of P-band informations indicated a good correlativity between the radio detection and ranging backscatter strength and the chief wood parametric quantities including bole biomass, tallness, age, diameter at chest tallness ( dbh ) , and radical country. Dobson et al. , ( 1992 ) showed an increasing scope of backscatter with altering biomass from C to P-band, every bit good as higher biomass degrees at which backscatter relationships to biomass saturate.
Hoekman, ( 1990 ) found hapless relationships between X- and C-band backscatter and volume and other base parametric quantities.The spaceborne systems, such as the Seasat SAR, SIR-B, SIR-C/X-SAR and ERS-1, ERS-2, JERS, ENVISAT-ASAR and late ALOS-PALSAR etc. were used for probes of boreal, temperature and sub-tropical forestry trial sites ( Ford et al. , 1988 ; Dobson et al. , 1992 ; Ranson et al. , 1995 ; Stofan et al. , 1995 ; Rignot et al.
, 1995 ) . These experiments and surveies have shown that radio detection and ranging is sensitive to forest structural parametric quantities including above-ground biomass ( Dobson et al. , 1992 ; Pulliainen et al. , 1994 ; Skriver et al. , 1994 ; Ferrazzoli et al. , 1995 ; Ranson et al. , 1996 ) .Kasischke et al.
, ( 1997 ) reviewed radio detection and ranging informations for ecological applications, including AGB appraisal. It is being reported in literature that the radio detection and ranging backscatter in the P and L sets is extremely correlated with major wood parametric quantities, such as tree age, tree tallness, DBH, radical country, and AGB. In peculiar, SAR L-Band informations have proven to be valuable for AGB appraisal ( Sader, 1987 ; Luckman et al. , 1997 ; Kurvonen et al.
, 1999 ; Sun et al. , 2002 ) . Kuplich et al. , ( 2000 ) used JERS-SAR informations for AGB appraisal of renewing woods and concluded that these informations had the possible to gauge AGB for immature, renewing woods. Luckman et al. , ( 1997 ) found that the longer-wavelength ( L-Band ) SAR image was more suited to know apart different degrees L-Band backscatter shows no sensitiveness to increased biomass denseness after a certain threshold, such as 100 dozenss ha-1, bespeaking that it is suited for gauging biomass of renewing woods in tropical parts.The radio detection and ranging backscattering coefficient is correlated with forest biomass and root volume ( Le Toan et Al.
1992 ; Israelsson et al. , 1994 ; Kasischke et al. , 1994, Dobson et al. , 1995 ) . The sensitiveness of Man-made Aperture Radar ( SAR ) information to forest root volume increases significantly as the radio detection and ranging wavelength additions ( Israelsson et al. , 1997 ) . The imaging procedure makes SAR suited for mapping parametric quantities related to forest biomass, like root volume ( Baker et al. , 1999 ; Israelsson et al.
, 1997 ; Pulliainen et al. , 1996 ) , entire turning stock ( Balzter et al. , 2000 ; Schmullius et al. , 1997 ) , LAI ( Imhoff et al. , 1997 ) , or above land net primary productiveness ( Bergen et al. , 1998 ) .The dependence of backscatter on above land biomass was observed and related to the incursion of the radiation into the canopy and interaction with the bole, where most of the volume, hence, biomass of the flora is concentrated ( Sader 1987, Le Toan et Al. 1992, Dobson et Al.
1992 ) . HV polarisation in longer wavelengths ( L or P set ) is the most sensitive to above land biomass ( Sader 1987, Le Toan et Al. 1992, Ranson et Al. 1997a ) because it originates chiefly from canopy volume dispersing ( Wang et al. 1995 ) , trunk sprinkling ( Le Toan et Al.
1992 ) and is less affected by the land surface ( Ranson and Sun 1994 ) .As forest backscatter in different wavelengths and polarisations originate from separate beds of a canopy, the usage of multiple channels or multi-step attacks ( e.g. , Dobson et Al.
1995 ) could be used to gauge entire above-ground biomass ( Kasischke et al. 1997 ) . Sun and Ranson ( 1994 ) estimated biomass in assorted conifer temperate forest upto 250 Mg/ha.Band ratios ( HH/HV and VV/VH ) were besides used for the above land biomass appraisal.
However, Dobson et Al. ( 1995 ) considered these band ratios excessively simplistic ( as the corresponding backscatter will be much higher for the few tall trees than for the many short 1s ) , although effectual in gauging biomass at higher scopes. In malice of this, a combination of sets and polarisations in a multi-step attack made possible the function of biomass in a assorted temperate forest upto 250 Mg/ha ( Dobson et al. 1995 ) . Establishing a strong nexus between backscatter and forest variables is an of import portion of the successful appraisal of forest biomass from backscatter. Models are frequently used to explicate the relationship between wood variables, dispersing mechanisms and SAR constellation parametric quantities ( Richards 1990, Kasischke and Christensen 1990 ) .
Another attack is the usage of statistical analysis, where forest variables are related to SAR backscatter by arrested development theoretical accounts ( Sader 1987, Le Toan et Al. 1992, Rauste et Al. 1994 ) . The combination of the two attacks, in most instances to measure the consequences of the predicted biomass or backscatter via arrested development ( Ranson and Sun 1994, Ferrazzoli et Al.
1997, Franson and Israelson 1999 ) . Statistical processs such as stepwise arrested development analysis were besides used to find the best set of sets and polarisations to know apart biomass degrees ( Ranson et al. 1997a ) .The three-band ( C, L, and P ) polarimetric AIRSAR detector has been used in many forest biomass surveies ( e.g. , Green, 1998 ; Kasischke et al.
, 1991, 1995 ; Moghaddam et al. , 1994 ; Ranson & A ; Sun, 1997 ) . The strongest correlativity between SAR backscatter and forest biomass has been reported in P-band and the weakest in C-band ( e.g. , Beaudoin et al.
, 1992 ; Dobson et al. , 1992 ; Israelsson et al. , 1992 ; Rauste et al. , 1992 ; Skriver & A ; Gudmandsen, 1992 ) . Studies utilizing dual-frequency ( C- and L-bands ) polarimetric SIR-C informations ( e.
g. , Dobson et al. , 1995 ; Harrell et al. , 1997 ) had shown the importance of L-band informations in cone-bearing forest biomass function.
The correlativity between forest biomass and the C-band backscatter measured by the ERS-1 and ERS-2 SAR detectors has been reported to be low in tropical woods ( Luckman et al. , 1997 ) . Kasischke et Al. ( 1994 ) described the C-band dynamic scope as low in immature loblolly pine woods due to biomass fluctuation. The usage of multi-temporal ERS dataset and a semi-empirical backscatter theoretical account for forest biomass appraisal has been discussed by Pulliainen et Al. ( 1996 ) . Repeat-pass interferometry with C-band ERS information has been studied for forest biomass function ( e.g.
, Fransson et al. , 2001 ; Luckman et al. , 2000 ) . Luckman et Al. ( 2000 ) obtained a coefficient of finding of 0.
805 between ERS-1/2 ( 1-day difference ) tandem coherency and the logarithm of forest biomass in renewing tropical woods ( biomass chiefly between 0 and 100 tons/ha ) .Even though L-band backscatter is significantly correlated with forest biomass, a confining factor is the impregnation of the backscatter-biomass relationship at some biomass degree ( Imhoff, 1993 ) . Imhoff ( 1995 ) reports the L-band impregnation degree at 40 tons/ha of dry biomass. Luckman et Al.
( 1998 ) found a impregnation of 60 tons/ha in a Brazilian trial site. In boreal woods, the 40 tons/ha bound corresponds to a forest root volume of about 70 m3/ha. Fransson and Israelsson ( 1999 ) obtained a impregnation degree of 143 m3/ha.
Kurvonen et Al. ( 1999 ) describe a impregnation degree of 225 m3/ha in two JERS SAR scenes. Hence, we can state that the impregnation degree may depend on the tree species and forest types every bit good as the land surface type. Kuplich et Al. ( 2000 ) found a strong L-band vs. biomass correlativity ( r =0.
77 ) in a Brazilian trial site ( re-growth after a clear cut ) while the correlativity was weak in a Cameronian trial site ( selective logging ) . Recently L-band ALOS PALSAR data become available and it is utile to look into their potency for land and forest screen function in different tellurian ecosystems. The aim of this probe was to analyze this freshly available data-set for wood screen function and above land biomass appraisal in the tropical wood parts.
VEGETATION CLASSIFICATION USING SAR DATA:
Recent reappraisals on the application of radio detection and ranging in forestry show that SAR systems have a good capableness in know aparting assorted types of ( tropical ) forest screen ( Van der Sanden, 1997 ; Varekamp, 2001 ; Qui & A ; ntilde ; 1s, 2002 ; Sgrenzaroli, 2004 ) .
In general categorization consequences are hapless if merely single-frequency, single-polarisation or individual flyover informations are used. In order to better the radio detection and ranging categorization capableness, either multi-temporal or multi-frequency informations are required. Multi-temporal informations, which may be acquired by airborne or satellite systems, are peculiarly of import to divide forest types ( Ferrazzoli et al. , 1999 ) . Multi-frequency or multi-polarisation informations are made available by several types of airborne SAR systems, such as the NASA/JPL AirSAR.
These are capable of geting interesting categorization consequences, due to features of multi-frequency interferometric and polarimetric characteristics, which can be associated to flora constructions.In recent old ages many research activities focused on the usage of SAR to analyze tropical rain forest. At Continental graduated table mosaics of all tropical rain forests have been created utilizing JERS-1 SAR images ( Siqueira et al. , 2000 ; Rosenqvist et al.
, 2000 ; Sgrenzaroli, 2004 ) and, for Africa, utilizing ERS-1 SAR ( De Grandi et al. , 1999 ) . At a larger graduated table research workers have focused their surveies on the development of inversion algorithms, cleavage and categorization techniques for polarimetric and interferometric SAR images and created a assortment of types of tropical rain forest categorizations, for illustration utilizing texture ( Van der Sanden, 1997 ; Oliver, 2000 ) and mapping single trees ( Varekamp, 2002 ) . Maping tropical rain forest types and its biophysical features with airborne SAR ( AirSAR ) was explored for the Amazon ( Hoekman and Quinones, 2002 ; Quinones, 2002 ) . Prakoso ( 2006 ) shows the latter type of application, utilizing AirSAR and TOPSAR airborne radio detection and ranging informations collected during the NASA PacRim-II run executed in 2000, was late studied for Kalimantan.Two chief attacks were used to sort SAR informations in the old surveies: ( 1 ) upper limit likelihood categorization ( MLE ) including supervised and unsupervised bunch analysis and ( 2 ) knowledge-based hierarchal determination trees ( Kasischke et al. 1997 ) . The extendibility of MLE categorization consequences to planetary graduated tables is normally impaired by the demand for localised preparation ( Kasischke et al.
1997 ) . Knowledge-based attacks have been proposed to get the better of this restriction by utilizing expressed relationships between backscatter and flora construction and so reclassification based on these links and floristic community ( Dobson et al. 1995, Kasischke et Al. 1997, Bergen et Al. 1998 ) . Maximum-a-posteriori ( MAP ) Bayesian classifier was developed for the categorization of multi- frequence polarimetric SAR information and differed to the MLE attack because of the alterations on the determination regulations about the category nature ( Saatchi and Rignot, 1997 ) .
Recent research has shown promising consequences utilizing cleavage methods ( Oliver 1998, Frery et Al. 1999, Grover et Al. 1999 ) . These methods consist of collection of pels with similar belongingss and bounds defined by the boundary lines of the sections ( Yanasse et al.
1997 ) .Artificial Neural Network ( ANN ) based classifiers are besides a promising attack. Among ANN advantages are facilitated incorporation of different types of informations which do non hold to suit any peculiar statistical distribution ( Atkinson and Tatnall, 1997 ) . The advantages of each attack depend on the suitableness of the categorization calculator to the available informations set, which will find a high truth on the categorization procedure. Field cognition is really important in sorting different types of wood utilizing SAR informations. Good field cognition, field informations and equal maps make algorithm preparation easier and accuracy appraisal of the concluding categorization.
Temperate and boreal forest types have been classified with radio detection and ranging informations ( Saatchi and Rignot, 1997 ; Bergen et Al. 1998, Williams et Al. 1999 ) . For direction stock list intents, nevertheless, radar information does non supply elaborate adequate information ( Leckie and Ranson, 1998 ) .
Nevertheless, radio detection and ranging informations can supply complementary information to aerial exposure ( Leckie and Ranson, 1998 ) and forest biophysical parametric quantities have been estimated ( Ranson and Sun, 1994 ; Dobson et Al. 1995, Ranson et Al. 1997 ) .When radio detection and ranging informations are combined with optical informations, forest function capablenesss are normally increased.
Manual reading is besides done utilizing SAR informations in tropical woods. Accurate automatic categorization of radio detection and ranging informations for tropical wood is still under development. Unifying categorization techniques ( Rignot et al. 1997 ) , the usage of calculators adapted to radar informations ( Nezry et al.
1993, Saatchi et Al. 1997, Saatchi et Al. 2000 ) and the usage of texture steps derived from SAR images ( Oliver 1998 ; Saatchi et Al. 2000 ) seem to be the tendencies for the high categorization truth of the flora on the Torrid Zones. The usage of the two frequence sets, C, L and/or P channels are indispensable to know apart between renewing forest and selectively logged forest ( vander Sanden and Hoekman, 1999 ) .
Unifying SAR informations with optical informations is normally done when analyzing the tropical woods ( Nezry et al. 1993, Rignot et Al. 1997, Araujo et Al. 1999 ) . Time series of optical images have been used as a mention in the field or anterior to field work to set up the age of glades and land cover history ( Foody et al. , 1997 ; Luckman et al. , 1997 ; Yanasse et al.
, 1997 ; Salas and Skole, 1998 ) . Much poorer categorization public presentations were observed in tropical wood categories.Balzter ( 2001 ) reviewed Interferometric Synthetic Aperture Radar ( InSAR ) for wood function and monitoring. Vegetation categorization is besides carried out utilizing spacial features of SAR informations i.e.
, texture steps. For backscatter, textural properties quantify the form of spacial fluctuations in the strength of backscatter ( Vander Sanden and Hoekman, 1999 ) . An optimised texture step depends on the statistical belongingss of the backscatter ( Oliver and Quegan, 1998 ) and is based on the statistical dependance between pels within a part ( Kurvonen and Hallikainen, 1999 ) . Local statistics texture steps are statistical minutes ( such as mean, lopsidedness, kurtosis and coefficient of fluctuation ( CV ) ) , of the window from which the texture of the image is extracted ( Kurvonen and Hallikainen, 1999 ) . Second-order texture steps ( such as information, energy, contrast, etc. ) relate to statistical dependance between pels in a given distance and way and are calculated from the grey-level accompaniment matrix ( GLCM ) ( Kurvonen and Hallikainen, 1999 ; Mather, 1999 ) .
Another attack for texture analysis includes the variogram that provides a concise description of the graduated table and form of spacial variableness in remotely sensed informations ( Curran et al. 1998 ) . While recent forest favoritism research has shown involvement on the textural attack, consequences are hard to generalize because of the assortment of physical environments and techniques used.
For temperate woods in Finland, texture steps ( co-efficient of fluctuation and four steps derived from GLCM ) from a multi-temporal set of SAR images were found to increase the truth of categorization consequences, even though the concluding truth was about 65 % ( Kurvonen and Hallikainen, 1999 ) . Use of a simple texture step ( such as the mean ) can ensue in accurate favoritism between forest types ( Yanasse et al. 1997 ; Podest and Saatchi, 1999 ) . Despite low truth in the favoritism of forest types, some writers report promoting representation of categories with typical texture signatures ( Miranda et al.
1998 ; Vander Sanden and Hoekman, 1999 ) . Possibly the spread in texture mold ( Oliver and Quegan, 1998 ) will be resolved with a better apprehension of the natural philosophies that governs backscatter and associated texture, given that texture is still a promising attack.The future space-borne SAR systems – particularly the C-band RISAT and L-band RISAT missions, will heighten the potency for broad country forest biomass function. The emerging Polarimetric interferometry technique can be used for forest biomass function.
As this technique requires to the full polarimetric informations acquisitions, the technique may turn out excessively demanding in footings of informations acquisition and country coverage. In to the full polarimetric manner, the SAR detectors of the close hereafter have an image swath breadth that is merely half of the nominal individual polarization swath. The so called dual-polarisation constellation, which includes coincident HH- and HV-polarised informations, covers the whole swath of the single-polarisation manner. Therefore, the usage of double polarisation, polarimetry and polarimetric interferometry is a cardinal research subject in the coming old ages.— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — – Man-made Aperture Radar gives response that is straight related to the sum of populating stuff with which it interacts. Radar backscatter ( signal reflected from the mark ) is relative to flora denseness up to a impregnation point that is dependent upon wavelength and polarisation of the radio detection and ranging. Beyond this impregnation point, farther addition in flora denseness either gives a decrease or no alteration in net backscatter due to extinction of the signal within the canopy bed. This consequence limits the capableness of given radio detection and ranging ( wavelength and polarisation ) to distinguish aboveground biomass degrees beyond the impregnation point.
The impregnation point can be extended through usage of multi-frequency and/or multi-polarization informations.Leaf optical belongingss are most straight expressed at the canopy degree in the NIR spectral part when foliage country ( LAI ) is really high. The major restriction of utilizing flora indices peculiarly NDVI is that they asymptotically approach a impregnation degree after a certain LAI 7.
0 is reached ( Thenkabali et al. , 2000 ) . However, the relationship depends on the dirt coefficient of reflection and NDVI saturates when the canopy reaches full coverage.
With their longer wavelength, microwaves are expected to perforate into the canopy and supply information about biomass ( foliage, subdivision, bole ) up to a higher impregnation degree compared to optical systems.