L-band SAR Research Articles

Theoretical accuracy of along-track displacement measurements from multiple-aperture interferometry (MAI).

The measurement of precise along-track displacements has been made with the multiple-aperture interferometry (MAI). The empirical accuracies of the MAI measurements are about 6.3 and 3.57 cm for ERS and ALOS data, respectively. However, the estimated empirical accuracies cannot be generalized to any interferometric pair because they largely depend on the processing parameters and coherence of the used SAR data. A theoretical formula is given to calculate an expected MAI measurement accuracy according to the system and processing parameters and interferometric coherence. In this paper, we have investigated the expected MAI measurement accuracy on the basis of the theoretical formula for the existing X-, C- and L-band satellite SAR systems. The similarity between the expected and empirical MAI measurement accuracies has been tested as well. The expected accuracies of about 2–3 cm and 3–4 cm (γ = 0.8) are calculated for the X- and L-band SAR systems, respectively. For the C-band systems, the expected accuracy of Radarsat-2 ultra-fine is about 3–4 cm and that of Sentinel-1 IW is about 27 cm (γ = 0.8). The results indicate that the expected MAI measurement accuracy of a given interferometric pair can be easily calculated by using the theoretical formula.

Analysis of L-band SAR backscatter and coherence for delineation of land-use/land-cover

In this study, we investigated the potential improvement of land-use/land-cover (LU/LC) classification using multidate backscatter intensity as well as interferometric coherence images derived from Advanced Land Observing Satellite phased array L-band synthetic aperture radar data. Four interferometric synthetic aperture radar data pairs in horizontal–horizontal polarizations were processed to obtain backscatter intensity and coherence images. From the analysis of these images, it was observed that backscatter values alone are not sufficient to separate certain LU/LC classes, e.g. forest and mining areas, due to similarities in the associated scattering mechanisms producing similar backscatter values. However, the temporal coherence values from these LU/LC features were found to be distinctly different. Supervised classifications using maximum-likelihood distance were performed with various combinations of data (three-date backscatter intensity and two-date backscatter intensity with corresponding coherence data) to generate LU/LC maps of the study area. The comparison of classification accuracies obtained for different combinations of data indicates that the classification accuracy is improved by adding coherence information to the backscatter intensity data compared to using the multidate backscatter intensity data alone. Thus, the analysis of backscatter intensity along with coherence is a better alternative than using backscatter intensity alone to improve the accuracy in LU/LC classification.

Equivalent Stacking of Polarimetric Synthetic Aperture Radar Interferograms Based on Analysis of Persistent and Distributed Scatterers

We propose a method of incorporating the copolarized observations (HH and VV observations) of multipolarimetric SAR acquisitions into persistent scatterer interferometry to estimate the surface displacement and the scatterer height. Interferograms obtained from HH and VV observations were equally weighted on the basis that the magnitude of the backscattered signals from persistent scatterers represented by odd-bounce scattering and even-bounce scattering are the same in the copolarized observations. Moreover, because of the coarser slant-range resolution of multipolarimetric SAR data, distributed scatterers (DSs) should be taken into account to obtain a spatial density of pixels that is sufficient to estimate the parameters of interest. To this aim, we applied a small baseline criterion to create the analyzed data sets, and we used space adaptive filtering to improve the signal-to-noise ratio of the DSs. We applied this method to estimate the ground subsidence rate in an area of groundwater extraction in Japan using the SAR images that are acquired by Advanced Land Observing Satellite/Phased Array Type L-band SAR (PALSAR) in the full polarimetric observation mode.

L-band ALOS PALSAR for biomass estimation of Matang Mangroves, Malaysia

This study has been carried out to evaluate the relationship between Advanced Land Observing Satellite (ALOS) Phased Array L-band SAR (PALSAR) backscattering coefficients and the aboveground biomass (AGB) of a managed mangrove forest in Malaysia. Matang Mangrove Forest Reserve known as Matang Mangroves was selected as the study area. It covers about 41,000ha of mangrove forest and is the largest single mangrove ecosystem in Peninsular Malaysia. A mosaic of L-band PALSAR fine beam dual (FBD) with 25mpixel spacing data for the year 2010 was provided by the Japan Aerospace Exploration Agency’s (JAXA) within the framework of the ALOS Kyoto and Carbon (K&C) Initiative. A total of 320 sampling plots that were collected in 2010 and 2011 were used in the study. The calculated plot-based AGB were correlated to the pixels/backscatter of PALSAR data. The best correlation function (i.e. from HV backscatter) was used to estimate and determine the aboveground biomass of the Matang Mangroves. The study found that the estimated AGB in Matang Mangroves ranged between 2.98 and 378.32±33.90Mgha−1 with an average of 99.40±33.90Mgha−1 and a total AGB of about 4.25 million Mg. The HV backscatter started to saturate at an AGB of 100Mgha−1 and the errors associated with the estimation occurred largely when the AGB exceeded 150Mgha−1. The study also found that the manipulation of polarisation was useful in discriminating succession levels of mangroves.

C- and L-band SAR interoperability: Filling the gaps in continuous forest cover mapping in Tasmania

This paper addresses the shortfall in L-band SAR data availability for the purpose of extracting spatially explicit information on forest cover, and the capacity to fill the gap using shorter wavelength C-band data. Specifically, comparisons are made of forest/non-forest (F/NF) extent derived from independent classification of ALOS PALSAR and RADARSAT-2 data acquired over Tasmania. Focussing on a temperate forest landscape, it was demonstrated that C-band SAR data can be used interchangeably with L-band data to produce a comparable estimate of forest/non-forest cover. Only partial interoperability is achieved however, given the limitations of dual polarisation C-band SAR in discerning forests of different growth stage and biomass.Ambiguities in F/NF status were more prevalent in the C-band classification, despite inclusion of topographic (surface elevation and slope) and textural features in the training dataset, with the key observations as follows: (i) Reduced dynamic range and greater overlap amongst F/NF classes; (ii) Similarities in volume scattering from harvested/regrowth eucalyptus areas and background native forest; (iii) Confusion between young pine plantation and harvested/regrowth areas due to comparable roughness; (iv) Less variation in backscatter and poorer separation of intact and managed eucalyptus forest; and (v) Reduced capacity for discrimination of forest types. In almost every case, the use of L-band data was preferable. The one exception was the limited separation of young pine plantation and harvested/regrowth areas in both C- and L-band data. A similar level of performance was achieved in the discrimination of mature pine plantation, and between young eucalyptus plantation and harvested/regrowth.The findings were restricted to single-date classification of C- and L- band data. The potential to extend a time-series of L-band observations over forest using dense time-series (i.e., intra-annual) C-band observations acquired in dual or quad polarisation mode, warrants further investigation. Where a positive trade-off exists between the benefits and costs of integrating these data, multi-frequency (e.g., C- and L-band) and multi-sensor approaches (e.g., SAR and optical) are a viable way forward for operational forest monitoring and carbon accounting.

Monitoring Everglades freshwater marsh water level using L-band synthetic aperture radar backscatter

The Florida Everglades plays a significant role in controlling floods, improving water quality, supporting ecosystems, and maintaining biodiversity in south Florida. Adaptive restoration and management of the Everglades requires the best information possible regarding wetland hydrology. We developed a new and innovative approach to quantify spatial and temporal variations in wetland water levels within the Everglades, Florida. We observed high correlations between water level measured at in situ gages and L-band SAR backscatter coefficients in the freshwater marsh, though C-band SAR backscatter has no close relationship with water level. Here we illustrate the complementarity of SAR backscatter coefficient differencing and interferometry (InSAR) for improved estimation of high spatial resolution water level variations in the Everglades. This technique has a certain limitation in applying to swamp forests with dense vegetation cover, but we conclude that this new method is promising in future applications to wetland hydrology research.

Operational performance of the ALOS global systematic acquisition strategy and observation plans for ALOS-2 PALSAR-2

With the launch of the Advanced Land Observing Satellite (ALOS) in 2006, the Japanese Space Agency (JAXA) took the initiative to implement the first global-scale systematic acquisition strategy for satellite sensors at fine and medium (2.5–20m) spatial resolution. Comprising all three sensors on ALOS (PALSAR, PRISM, AVNIR-2), the plan was designed to serve all ALOS user categories and aimed at producing spatially and temporally consistent baseline coverages over the planet on a repetitive basis, to accommodate systematic global-scale, fine-resolution, monitoring of the environment. Unlike the common background missions defined for most fine-resolution Earth Observation satellites, the observation strategy was implemented as a top-level foreground mission with a priority second only to that of special observation requests and emergency observations and sensor calibration.While the ALOS mission regrettably ended in April 2011, the global acquisition strategy nevertheless produced a comprehensive and homogeneous global archive in which consistent time-series of data are available for any arbitrary land area on Earth (excluding Antarctica >77.5° South latitude, which could not be reached by the sensors). Clouds and haze inevitably constituted limitations for the optical sensors, while for the PALSAR instrument, two cloud-free and near-gap free (~95%) global coverages were achieved annually during the 4.5years in operations. Previously, such uniform data archives existed only for coarse-resolution sensors such as AVHRR, MERIS and MODIS. The ALOS BOS supported a variety applications from local to global scales, ranging from structural deformation, monitoring of wetlands regional inundation patterns and mapping of forest extent and changes over nations and continents at spatial resolutions as fine as 10m.The Advanced Land Observing Satellite 2 (ALOS-2) was launched on May 24, 2014. Equipped with an enhanced L-band SAR sensor (PALSAR-2), ALOS-2 resumes the global wall-to-wall acquisitions to assure continuity and consistency with JAXA’s global mission objectives and unique L-band SAR archive created by ALOS PALSAR.

Correction of Atmospheric Phase Screen in Time Series InSAR Using WRF Model for Monitoring Volcanic Activities

One of the main limitations in measuring ground deformation using synthetic aperture radar interferometry (InSAR) is atmospheric phase delay effects. In volcanic regions, the atmospheric phase delay effects can cause serious problems in detecting volcanic unrest because atmospheric thickness is inversely related with the elevation of a volcanic mountain. It is commonly known that the atmospheric phase screen (APS) can be decomposed spatially into stratified and turbulent components. In this paper, the stratified and turbulent atmospheric conditions of a volcanic area were simulated using weather research and forecasting (WRF) model, and the simulated atmospheric conditions were compared with in situ radiosonde data. The comparison results proved that the stratified APS from the WRF model could reflect the reasonable patterns of seasonal changes and vertical profiles with dependable quality. We also found that the stratified APS was significantly correlated with time and sometimes severely contaminated the quality of volcanic deformation estimation. These results indicate that the temporal high-pass (HP) filtering, which has been usually applied in time series InSAR analysis for extracting and removing APS, cannot work properly in volcanic area. Thus, we propose a new method that employs the stratified APS obtained from the WRF model by correlating with topography and the (residual) turbulent APS that can be effectively eliminated by temporal HP filtering in persistent scatterer interferometry (PSI). We applied the proposed method (atmosphere-corrected PSI) to Advanced Land Observing Satellite Phased Array type L-band SAR data that cover Shinmoedake volcano, Japan, and found that the estimated surface deformation and APS agreed well with those measured from GPS and Moderate-Resolution Imaging Spectroradiometer data, respectively.

Classifying Complex Mountainous Forests with L-Band SAR and Landsat Data Integration: A Comparison among Different Machine Learning Methods in the Hyrcanian Forest

Forest environment classification in mountain regions based on single-sensor remote sensing approaches is hindered by forest complexity and topographic effects. Temperate broadleaf forests in western Asia such as the Hyrcanian forest in northern Iran have already suffered from intense anthropogenic activities. In those regions, forests mainly extend in rough terrain and comprise different stand structures, which are difficult to discriminate. This paper explores the joint analysis of Landsat7/ETM+, L-band SAR and their derived parameters and the effect of terrain corrections to overcome the challenges of discriminating forest stand age classes in mountain regions. We also verified the performances of three machine learning methods which have recently shown promising results using multisource data; support vector machines (SVM), neural networks (NN), random forest (RF) and one traditional classifier (i.e., maximum likelihood classification (MLC)) as a benchmark. The non-topographically corrected ETM+ data failed to differentiate among different forest stand age classes (average classification accuracy (OA) = 65%). This confirms the need to reduce relief effects prior data classification in mountain regions. SAR backscattering alone cannot properly differentiate among different forest stand age classes (OA = 62%). However, textures and PolSAR features are very efficient for the separation of forest classes (OA = 82%). The highest classification accuracy was achieved by the joint usage of SAR and ETM+ (OA = 86%). However, this shows a slight improvement compared to the ETM+ classification (OA = 84%). The machine learning classifiers proved t o be more robust and accurate compared to MLC. SVM and RF statistically produced better classification results than NN in the exploitation of the considered multi-source data.

Forest Height Inversion Using Dual-pol Polarimetric SAR Interferometry

Polarimetric Synthetic Aperture Radar Interferometry (PolInSAR) has been extensively applied for forest parameter inversion over different frequencies and polarimetric conditions. So far, most research was based on full-pol SAR images with relatively small coverage. A spaceborne SAR system will have the potential for PolInSAR applications used for global forest monitoring. Spaceborne dual-pol SAR images usually have higher resolution and larger swath than full-pol mode. In this paper, forest height retrieval was attempted by PolInSAR from a L-band spaceborne dual-pol SAR pairs using HH and HV channels. The random volume over ground (RVoG) model was used to retrieve the height and the coherence optimization method was extended to the dual-pol PolInSAR, which makes use of polarimetry to enhance the quality of SAR interferograms. The three-stage process is also used in the dual-pol PolInSAR technique. Finally, the experimental test was performed for forest height estimation on the dual-pol L-band SAR data of the Saihanba forest acquired by the ALOS PALSAR sensor in 2009.

Swell parameters retrieval using ALOS/PALSAR data and its application

We have developed retrieval methods for swell parameters such as wave height, wavelength, and wave propagation direction from images of PALSAR on board ALOS taken over coastal seas around Japan. The retrieval methods for wavelength and wave direction use the PALSAR image spectrum. The SAR image spectrum has an inherent 180° direction ambiguity, which can be solved using the fact that the swells only propagate toward the coast. Using the coupling swell spectra and the PALSAR image spectra, an empirical L-band modulation transfer function was obtained for the wave height retrieval through linear regression analysis. A comparison between SAR-derived and in situ wave parameters showed close agreement. For the wavelength and wave direction, the biases were 10.4 m and 1.3° the root mean-square errors (RMSE) were 18.3 m and 15.5° and the correlation coefficients were 0.93 and 0.94, respectively. These results demonstrate that the L-band SAR imaging mechanism for swells is mostly linear. For the wave height, the bias was 0.08 m, the RMSE was 0.30 m, and the correlation coefficient was 0.80. The coastal water depth was calculated from the retrieved wavelength. The comparison shows that the result is better when the water depth is less than 50 meters.

Impact of Tree Species on Magnitude of PALSAR Interferometric Coherence over Siberian Forest at Frozen and Unfrozen Conditions

Numerous studies demonstrated the potential of the magnitude of interferometric coherence |γ| for forest growing stock volume (GSV) estimation in boreal forests. Coherence derived from images acquired under frozen conditions proved to be of specific interest. This also applies to PALSAR coherence, although affected by a comparatively large temporal baseline of at least 46 days. However, when working with spaceborne L-band data, acquired under unfrozen conditions, a large spread of |γ| was observed at all GSV levels. This scatter negatively affects the correlation of GSV and |γ|. So far, the impact of tree species on |γ| has rarely been studied in this context, although the different tree geometries are likely to have an impact on volumetric decorrelation. This paper presents the results of a study investigating the impact of tree species on PALSAR coherence employing 36 interferograms. The observations show only a small impact of the tree species on |γ| during frozen conditions. At unfrozen conditions, the impact is about three times larger. Deciduous species (aspen, birch, larch) exhibit the lowest |γ|, while coniferous species (fir, pine) feature the highest |γ|. For example, at unfrozen conditions, the |γ| of fir is 0.15 greater than the |γ| of larch, while the mean |γ| of dense forest is 0.38. Accordingly, the impact of tree species on |γ| under unfrozen conditions causes a portion of the observed spread of the GSV-|γ| relationship. Consequently, when aiming at |γ| based GSV assessment using L-band SAR data acquired during unfrozen conditions, the impact of the species on |γ| needs to be considered. For studies aiming at |γ| based GSV estimation across species, PALSAR data acquired at frozen conditions is preferable.

Oil spill detection from SAR image using SVM based classification

Abstract. In this paper, the potential of fully polarimetric L-band SAR data for detecting sea oil spills is investigated using polarimetric decompositions and texture analysis based on SVM classifier. First, power and magnitude measurements of HH and VV polarization modes and, Pauli, Freeman and Krogager decompositions are computed and applied in SVM classifier. Texture analysis is used for identification using SVM method. The texture features i.e. Mean, Variance, Contrast and Dissimilarity from them are then extracted. Experiments are conducted on full polarimetric SAR data acquired from PALSAR sensor of ALOS satellite on August 25, 2006. An accuracy assessment indicated overall accuracy of 78.92% and 96.46% for the power measurement of the VV polarization and the Krogager decomposition respectively in first step. But by use of texture analysis the results are improved to 96.44% and 96.65% quality for mean of power and magnitude measurements of HH and VV polarizations and the Krogager decomposition. Results show that the Krogager polarimetric decomposition method has the satisfying result for detection of sea oil spill on the sea surface and the texture analysis presents the good results.

Retrieval of tropical forest biomass information from ALOS PALSAR data

Advanced Land Observing Satellite Phased Array L-band Synthetic Aperture Radar (ALOS PALSAR) data from different observation modes were analysed to determine (1) which observation mode most accurately retrieves tropical forest biomass information and (2) whether different modes, when considered together, yield improved results in comparison to identical data-sets analysed independently. We performed regression analysis to estimate above-ground forest biomass using PALSAR backscatter data for natural and planted forests in south-eastern Bangladesh. The coefficient of determination (r 2) was lower or equal to 0.499 (n = 70) when PALSAR data from different observation modes were separately considered, but increased sharply when one class (rubber) is dropped and average backscatter of fine beam single (FBS) and polrimetric (PLR) modes are used in the analysis. The results presented in this article are useful for both regional and global forest biomass inventories and fixing acquisition modes for planned L-band SAR missions.

Detecting changes in surface soil moisture content using differential SAR interferometry

The feasibility of measuring changes in surface soil moisture content with differential interferometric synthetic aperture radar (DInSAR) has received little attention in comparison with other active microwave techniques. In this study, multi-polarization C- and L-band DInSAR is explored as a potential tool for the measurement of changes in surface soil moisture in agricultural areas. Using 10 ascending phased array L-band SAR (PALSAR) scenes acquired by the Japanese Advanced Land Observing Satellite (ALOS) and 12 descending advanced SAR (ASAR) scenes acquired by the European ENVISAT satellite between July 2007 and November 2009, a series of 27 differential interferograms covering a common study area over southern Ireland were generated to investigate whether small-scale changes in phase are linked to measured soil moisture changes. Comparisons of observed mean surface displacement and in situ mean soil moisture change show that C-band cross-polarization pairs displayed the highest correlation coefficients over both the barley (correlation coefficient, r = 0.51, p = 0.04)- and potato crop (r = 0.81, p = 0.003)-covered fields. Current results support the hypothesis that a soil moisture phase contribution exists within differential interferograms covering agricultural areas.

Sea Ice Detection in the Sea of Okhotsk Using PALSAR and MODIS Data

In this paper, we propose a new method for improving detection of sea ice through use of PALSAR (Phased-Array type L-band SAR) polarimetric data. Unlike traditional methods that are based on backscattering coefficient, our proposed method utilizes scattering entropy to detect sea ice. We tested this method by comparing sea ice area derived from PALSAR fully polarimetric data covering the Sea of Okhotsk acquired in 2009 and 2010 to that derived from reference MODIS (Moderate Resolution Imaging Spectroradiometer) data of the same region. We applied discriminant analysis to samples of sea ice and open water to determine the threshold of PALSAR backscattering coefficient and scattering entropy needed to discriminate sea ice from open water. We found that sea ice area derived from PALSAR data was equivalent to that derived from MODIS data, suggesting that our proposed method was reliable in the detection of sea ice in the Sea of Okhotsk.

Full polarimetric PALSAR-based land cover monitoring in Cambodia for implementation of REDD policies

Forest cover monitoring plays an important role in the implementation of climate change mitigation policies such as Kyoto protocol and Reducing Emissions from Deforestation and Forest Degradation (REDD). In this study, we have monitored land cover using the PALSAR (Phased Array type L-band Synthetic Aperture Radar) full polarimetric data based on incoherent target decomposition. Supervised classification technique has been applied on Cloude–Pottier decomposition, Freeman–Durden three component, and Yamaguchi four component decomposition for accurate mapping of different types of land cover classes. Based on confusion matrix derived from the predicted and defined pixels, the evergreen and sparsely deciduous forests have shown high producer's accuracy by Freeman–Durden three component and Yamaguchi four component classifications. The overall accuracy of Maximum Likelihood Classification by Yamaguchi four component is 94.1% with 0.93 kappa coefficient as compared to the 90.3% with 0.88 kappa coefficient by Freeman–Durden three component and 89.7% with 0.88 kappa coefficient by Cloude–Pottier decomposition. High accuracy of classification in a forested area using full polarimetric PALSAR data may have been because of high penetration of L-band SAR. The content of this study could be useful for the forest cover mapping during cloudy days needed for proper implementation of REDD policies in Cambodia.

Extraction of Underwater Laver Cultivation Nets by SAR Polarimetric Entropy

This letter describes a technique of extracting and estimating the underwater laver cultivation nets by using the entropy analysis of polarimetric synthetic aperture radar (PolSAR) data. The cultivation nets are placed at 10-20 cm below the sea surface, so that the Bragg waves responsible for L-band radar backscatter do not fully develop in this area of effectively shallow water. Consequently, the surface becomes smooth, and the backscatter radar cross section (RCS) becomes small in comparison with that from deep water without cultivation nets. If RCS from the cultivation area is at the system noise level, the image can be considered as arising from a random process, and the polarimetric entropy should be higher than the open sea area where the radar backscatter is dominated by the single-bounce surface scattering process. We will show that, using the data acquired by Phased Array L-band SAR onboard the Advanced Land Observing Satellite over the Tokyo Bay, Japan, the polarimetric entropy is an effective means of extracting underwater cultivation areas in comparison with the amplitude images. The area of the laver cultivation is then estimated by applying a constant false alarm rate to the entropy images to yield good agreement with the ground-truth data.

Stand-Level Stem Volume of Boreal Forests From Spaceborne SAR Imagery at L-Band

This paper presents a modified robust stem volume retrieval approach suitable for use with L-band SAR imagery. Multitemporal dual-polarization SAR imagery acquired by ALOS PALSAR during the summer-autumn 2007 is used in the study, along with stand-wise forest inventory data from two boreal forest sites situated in central Finland. The average sizes of forest stands at the study sites were 3 ha and 4.8 ha. The method used employs model fitting with an inverted semi-empirical boreal forest model, and takes advantage of the multitemporal aspect in order to improve the stability and accuracy of stem volume estimation. Multitemporal combination of model output in a multivariate regression framework allows volume estimates to be obtained with an RMSE about 43% of the mean of 110 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /ha, and a coefficient of determination R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.71 in the best case. The methodology used can be employed to produce large-area stem volume maps from dual-polarization ALOS PALSAR imagery mosaics.

Advanced Land Observing Satellite Phased Array Type L-Band SAR (ALOS PALSAR) to Inform the Conservation of Mangroves: Sundarbans as a Case Study

Mangroves are an important bulkhead against climate change: they afford protection for coastal areas from tidal waves and cyclones, and are among the most carbon-rich forests in the tropics. As such, protection of mangroves is an urgent priority. This work provides some new information on patterns of degradation in the Sundarbans, the largest contiguous mangrove forest in the world, which are home to more than 35 reptile species, 120 commercial fish species, 300 bird species and 32 mammal species. Using radar imagery, we contrast and quantify the recent impacts of cyclone Sidr and anthropogenic degradation on this ecosystem. Our results, inferred from changes in radar backscatter, confirm already reported trends in coastline retreat for this region, with areas losing as much as 200 m of coast per year. They also suggest rapid changes in mangrove dynamics for Bangladesh and India, highlighting an overall decrease in mangrove health in the east side of the Sundarbans, and an overall increase in this parameter for the west side of the Sundarbans. As global environmental change takes its toll in this part of the world, more detailed, regular information on mangroves’ distribution and health is required: our study illustrates how different threats experienced by mangroves can be detected and mapped using radar-based information, to guide management action.