Polarimetric Radar Backscatter Research Articles

Estimation of lunar surface roughness using Chandrayaan-2 full-polarimetric DFSAR data

Polarimetric radar backscattering coefficients depend on the impinging frequency, dielectric constant, incidence angle, polarization, and surface roughness. Therefore, surface roughness can be estimated using high-resolution polarimetric radar datasets. For this purpose, we examine the utility of model-based and eigenvalue-based decomposition approaches in this work. While using model-based decomposition, the dielectric constant is estimated at the outset. Thereafter, the rms height is derived from a scalar multiplier that models surface roughness. We also propose a modified single-bounce eigenvalue relative difference (SERD) and establish that it is a better indicator of surface roughness than the circular polarization coherence and the original SERD for non-reflection-symmetric lunar terrain. The Chandrayaan-2 Dual Frequency Synthetic Aperture Radar (DFSAR) datasets acquired over three Apollo mission landing sites are used for demonstration.

Radiometric Terrain Flattening of Geocoded Stacks of SAR Imagery

We have described an efficient approach to radiometrically flatten geocoded stacks of calibrated synthetic aperture radar (SAR) data for terrain-related effects. We have used simulation to demonstrate that, for the Sentinel-1 mission, one static radiometric terrain-flattening factor derived from actual SAR imaging metadata per imaging geometry is sufficient for flattening interferometrically compliant stacks of SAR data. We have quantified the loss of precision due to the application of static flattening factors, and show that these are well below the stated requirements of change-detection algorithms. Finally, we have discussed the implications of applying radiometric terrain flattening to geocoded SAR data instead of the traditional approach of flattening data provided in the original SAR image geometry. The proposed approach allows for efficient and consistent generation of five different Committee of Earth-Observation Satellites (CEOS) Analysis-Ready Dataset (ARD) families—Geocoded Single-Look Complex (GSLC), Interferometric Radar (InSAR), Normalized Radar Backscatter (NRB), Polarimetric Radar (POL) and Ocean Radar Backscatter (ORB) from SAR missions in a common framework.

Permafrost Dynamics Observatory (PDO): 2. Joint Retrieval of Permafrost Active Layer Thickness and Soil Moisture From L‐Band InSAR and P‐Band PolSAR

AbstractSeasonal subsidence induced by ground ice melt can be measured by interferometric synthetic aperture radar (InSAR) techniques to infer active layer thickness (ALT) in permafrost regions. The magnitude of subsidence depends on both how deep the soil thawed and how much ice/water content existed in the active layer soil. To provide the later, P‐band polarimetric synthetic aperture radar (PolSAR) backscatter is used due to its sensitivity to subsurface soil moisture and freeze/thaw conditions. In this study, which is the second in a two‐part series of Permafrost Dynamics Observatory (PDO), we exploit L‐band InSAR subsidence and P‐band PolSAR backscatter in a joint retrieval scheme to simultaneously estimate ALT and soil moisture profile of permafrost active layer. Both subsidence and backscatter are explicitly characterized by physics‐based models and share a common set of soil parameters including porosity and water saturation profiles. The PDO joint retrieval has been applied to the L‐ and P‐band SAR data acquired by National Aeronautics and Space Administration/Jet Propulsion Laboratory's Uninhabited Aerial Vehicle Synthetic Aperture Radar over Alaska and western Canada during the 2017 Arctic‐Boreal Vulnerability Experiment (ABoVE) airborne campaign. This high‐resolution (30 m) regional estimates of ALT and soil moisture profile spanning over the ABoVE study domain can help link the ground‐based field surveys with satellite observations to further understand the permafrost and active layer soil process dynamics to disturbances and climate change occurring across the northern circumpolar region.

Machine learning algorithms for soil moisture estimation using Sentinel-1: Model development and implementation

The present study provided the first-time comprehensive evaluation of 12 advanced statistical and machine learning (ML) algorithms for the Soil Moisture (SM) estimation from dual polarimetric Sentinel-1 radar backscatter. The ML algorithms namely support vector machine (SVM) with linear, polynomial, radial and sigmoid kernel, random forest (RF), multi-layer perceptron (MLP), radial basis function (RBF), Wang and Mendel’s (WM), subtractive clustering (SBC), adaptive neuro fuzzy inference system (ANFIS), hybrid fuzzy interference system (HyFIS), and dynamic evolving neural fuzzy inference system (DENFIS) were used. Extensive field samplings were performed for collection of in-situ SM data and other parameters from the selected sites for seven different dates and at two different locations (Varanasi and Guntur District, India), concurrent to Sentinel-1 overpasses. The backscattering coefficients were considered as input variables and SM as output variable for the training, validation and testing of the ML algorithms. The site at Varanasi was used for the training, validation and testing of the models. On the other hand, the Guntur site was used as an independent site for checking the model performance, before finalizing the algorithms. The performances of different trained algorithms were evaluated in terms of correlation coefficient ( r ), root mean square error (RMSE) (in m 3 /m 3 ) and bias (in m 3 /m 3 ). The study identified the RF, SBC and ANFIS as the top three best performing models with comparable and promising SM estimation. In order to test the robustness of these best models (RF, SBC and ANFIS), further performance analysis was performed to the independent datasets of the Varanasi and Guntur test sites, which indicates that the performance of these three models were consistent and SBC can be recommended as the best among all for SM estimation.

Surface-based Ku- and Ka-band polarimetric radar for sea ice studies

Abstract. To improve our understanding of how snow properties influence sea ice thickness retrievals from presently operational and upcoming satellite radar altimeter missions, as well as to investigate the potential for combining dual frequencies to simultaneously map snow depth and sea ice thickness, a new, surface-based, fully polarimetric Ku- and Ka-band radar (KuKa radar) was built and deployed during the 2019–2020 year-long MOSAiC international Arctic drift expedition. This instrument, built to operate both as an altimeter (stare mode) and as a scatterometer (scan mode), provided the first in situ Ku- and Ka-band dual-frequency radar observations from autumn freeze-up through midwinter and covering newly formed ice in leads and first-year and second-year ice floes. Data gathered in the altimeter mode will be used to investigate the potential for estimating snow depth as the difference between dominant radar scattering horizons in the Ka- and Ku-band data. In the scatterometer mode, the Ku- and Ka-band radars operated under a wide range of azimuth and incidence angles, continuously assessing changes in the polarimetric radar backscatter and derived polarimetric parameters, as snow properties varied under varying atmospheric conditions. These observations allow for characterizing radar backscatter responses to changes in atmospheric and surface geophysical conditions. In this paper, we describe the KuKa radar, illustrate examples of its data and demonstrate their potential for these investigations.

Retrieving soil moisture profiles based on multifrequency polarimetric radar backscattering observations. Theoretical case study

ABSTRACT In this theoretical work, a dual-frequency polarimetric method is proposed for measuring moisture profiles in the topsoil up to 0.30 m thick. A case of measuring soil moisture profiles, which monotonically changes with depth, during 37 days after irrigation is considered. Original values of backscattering coefficients are calculated by the Oh model and by the small perturbation method at frequencies of 5.4 GHz and 435 MHz, respectively. In these calculations, we used measured moisture profiles and spectroscopic refractive mixing dielectric model of non-saline mineral soil with a clay fraction of 9.1%. Soil moisture profiles are retrieved by solving the inverse problem, the cost function of which is constructed based on the co- and cross-polarized ratios, calculated at two frequencies for the measured and modelled soil moisture profiles. An exponential function is used as a modelled soil moisture profile. It is shown that the standard deviation between the retrieved and measured soil moisture values in the surface layer 0.30 m thick appears to be ≤0.02 m3 m−3 (theoretical limit), and the determination coefficient is 0.881. The study shows a promising path towards developing multi-frequency radar systems for remote sensing of soil moisture profiles using satellites-based and unmanned aerial vehicles air-based platforms.

The backscattering contribution of soybean pods at L-band

L-band (1.25 GHz) radar measurements of a soybean canopy indicate that the emergence of seed pods is a significant contributor to the backscatter during the late stages of the growing season. In order to validate the measured data, a realistic scattering model of the soybean canopy is developed. The parameters of the soybean canopy and underlying soil used in the model vary over the growing season based on in situ measurements. Scattering amplitudes for soybean leaves are modeled analytically by using a thin disk approximation; stem and pods are jointly modeled using a numerical electromagnetic field solver. These scattering amplitudes are together incorporated into a coherent scattering model to obtain the backscattering coefficient for VV- and HH-polarizations. The modeling results show good agreement with the radar field measurements, having RMSEs of 0.51 dB for VV-pol and 1.1 dB for HH-pol. Both measured data and modeled results show that the change of soil moisture can be accurately monitored by L-band backscatter. It is also found that the difference between HH- and VV-polarized backscatter increases as the size of the soybean pods becomes larger. A method is developed here to estimate the number of pods in a soybean canopy based on polarimetric radar backscatter at L-band.

Adaptive Laplacian Eigenmap-Based Dimension Reduction for Ocean Target Discrimination

It is well known that polarimetric synthetic aperture radar (PolSAR) backscattering features are highly influenced by the variation of incidence angle (VIA), which usually hampers the classification of most grazing-angle-sensitive targets, such as land and ocean targets. To relieve this issue, various feature extraction approaches have been suggested to enhance the class discriminability while reducing the observed feature dimensionality. The Laplacian eigenmap-based dimension reduction (DR) has been proven to be an effective way to deal with VIA problems, provided that the manifold parameters [e.g., the heat kernel (HK)] have been optimally sought, which is often difficult in practice. In this letter, an adaptive Laplacian eigenmap-based DR method is presented to find a learned subspace where the local geometry with discriminative prior knowledge is preserved as much as possible while near optimal HK and scale factor parameters are automatically identified. The learned feature representation is then employed for the subsequent classification. The improved Laplacian eigenmap algorithm was validated by three uninhabited-aerial-vehicle-synthetic-aperture-radar L-band PolSAR images from the Gulf Deepwater Horizon oil spill, which were clearly impacted by the VIA phenomenon. The experimental results showed that the proposed algorithm works well in ocean target discrimination compared with the current common methods.

Experimental Characterization of Polarimetric Radar Backscatter Response of Distributed Targets at High Millimeter-Wave Frequencies

Subterahertz frequencies between 100 and 300 GHz remain an untapped portion of the frequency spectrum for many radar- and radiometer-based remote sensing applications. This can be attributed in part to the lack of knowledge of the phenomenology of signal interaction with terrain at these frequencies. This paper examines recently acquired polarimetric radar backscatter data of different types of surfaces using a newly constructed polarimetric instrumentation radar operating at 222 GHz. At this frequency bare surfaces such as asphalt and dirt are electrically rough with subsurface aggregate sizes comparable with the wavelength resulting in substantial volume and surface scattering. The data show strong backscatter responses from bare surfaces with angular dependence proportional to the cosine square of the incidence angle along with significant depolarization (between −8 and −4 dB). The data for vegetation-covered surfaces show weak dependence on incidence angle and appreciable depolarization (between −12 and −6 dB). Empirical models for bare and vegetation-covered surfaces are proposed.

Surface roughness and breaking wave properties retrieved from polarimetric microwave radar backscattering

AbstractOcean surface roughness and wave breaking are the two main contributors of radar backscattering from the ocean surface. The relative weightings of the two contributions vary with the microwave polarization: the VV (vertical transmit vertical receive) is dominated by the Bragg resonance scattering mechanism, and the HH (horizontal transmit horizontal receive) and VH (horizontal transmit vertical receive or vertical transmit horizontal receive) contain nontrivial non‐Bragg contributions mainly produced by breaking features. A method is developed to obtain the short‐scale properties of ocean surface roughness and wave breaking from Ku, C, and L band polarimetric sea returns. The results are used for quantitative evaluation of the ocean surface roughness spectral models and for deriving understanding of the breaking contribution important to microwave ocean remote sensing, in particular its dependence on wind speed, microwave frequency, and incidence angle. Implications of the results to air‐sea interaction applications are discussed.

An Error Model for Biomass Estimates Derived From Polarimetric Radar Backscatter

Estimating the amount of above ground biomass in forested areas and the measurement of carbon flux through the quantification of disturbance and regrowth are critical to develop a better understanding of ecosystem processes. Well-resolved and globally consistent inventories of forest carbon must rely on remote sensing measurements, particularly from polarimetric radars. While a wide variety of studies conducted over the past three decades have shown how radar polarimetric measurements can be used to estimate above ground carbon for regions with less than 100 Mg of biomass per hectare, there is no established methodology for assessing biomass estimation accuracy based on a priori instrument and mission parameters. In this paper, a framework for assessing biomass estimation accuracy is presented that is a blend of the basic imaging physics and empirically derived parameters that describe various relationships between biomass and radar polarimetric observable quantities. The implications of this error model on the design and performance of a polarimetric radar are explored using instrument, mission, and science parameters from a notional Earth observing mission.

Study on Moisture Effects on Polarimetric Radar Backscatter from Forested Terrain

Study on Moisture Effects on Polarimetric Radar Backscatter from Forested Terrain

Basal Area and Biomass Estimates of Loblolly Pine Stands Using L-band UAVSAR

Fully polarimetric L-band Synthetic Aperture Radar ( SAR ) backscatter was collected using NASA ’s Unmanned Aerial Vehicle ( UAV ) SAR and regressed with in situ measurements of basal area ( BA ) and above ground biomass ( AGB ) of mature loblolly pine stands in North Carolina. Results found HH polarization consistently displayed the lowest correlations where HV and VV exhibited the highest correlations in all groups for both BA and AGB . When plantation stands were analyzed separately (plantation versus natural), correlation improved signifi cantly for both BA (R 2 = 0.65, HV ) and AGB (R 2 = 0.66, VV ). Similarly, results improved when natural stands were analyzed separately resulting in the highest correlation for AGB (R 2 = 0.63, HV and VV ). Data decomposition using the Freeman 3-component model indicated that the relative low correlations were due to the saturation of the L-band backscatter across the majority of the study area.

Computation of Radar Scattering From Heterogeneous Rough Soil Using the Finite-Element Method

A 2-D vector-element-based finite-element method (FEM) is used to calculate the radar backscatter from 1-D bare rough soil surfaces which can have an underlying heterogeneous substrate. Monte Carlo simulation results are presented for scattering at L-band (λ = 0.24 m). For homogeneous soils with rough surfaces, the results of FEM are compared with the predictions of the small perturbation method. In the case of heterogeneous substrates, soil moisture (and, hence, soil permittivity) is assumed to vary as a function of depth. In this case, the results of FEM are compared with those of the transfer matrix method for flat soil surfaces. In both cases, a good agreement is found. For homogeneous rough soils, it is found that polarimetric radar backscatter and copolarized phase difference have a nonlinear relationship with soil moisture. Finally, it is found that the nature of the soil moisture variation in the top few centimeters of the soil has a strong influence on the backscatter and, hence, on the inferred soil moisture content.

Regression-Based Retrieval of Boreal Forest Biomass in Sloping Terrain Using P-Band SAR Backscatter Intensity Data

A new biomass retrieval model for boreal forest using polarimetric P-band synthetic aperture radar (SAR) backscatter is presented. The model is based on two main SAR quantities: the HV backscatter and the HH/VV backscatter ratio. It also includes a topographic correction based on the ground slope. The model is developed from analysis of stand-wise data from two airborne P-band SAR campaigns: BioSAR 2007 (test site: Remningstorp, southern Sweden, biomass range: 10–287 tons/ha, slope range: 0–4 $^{\circ}$ ) and BioSAR 2008 (test site: Krycklan, northern Sweden, biomass range: 8–257 tons/ha, slope range: 0–19 $^{\circ}$ ). The new model is compared to five other models in a set of tests to evaluate its performance in different conditions. All models are first tested on data sets from Remningstorp with different moisture conditions, acquired during three periods in the spring of 2007. Thereafter, the models are tested in topographic terrain using SAR data acquired for different flight headings in Krycklan. The models are also evaluated across sites, i.e., training on one site followed by validation on the other site. Using the new model with parameters estimated on Krycklan data, biomass in Remningstorp is retrieved with RMSE of 40–59 tons/ha, or 22–33% of the mean biomass, which is lower compared to the other models. In the inverse scenario, the examined site is not well represented in the training data set, and the results are therefore not conclusive.

ALOS PALSAR L-band polarimetric SAR data and in situ measurements for leaf area index assessment

Leaf area index (LAI), a key parameter controlling crop growth and yield models, has been widely estimated using optical satellite measurements. The estimation of LAI from high-resolution optical satellite data is limited by cloudy conditions and this may be a problem when systematic monitoring during the growing season is required. Synthetic Aperture Radar data are less susceptible to atmospheric effects than optical data and have been related to standing biomass over a number of landscapes. Here we quantify the relationship between LAI and both Advanced Land Observing Satellite (ALOS) Phased Array Synthetic Aperture Radar (PALSAR) L-band data and ENVISAT Advanced Synthetic Aperture Radar C-band data under relatively uniform soil moisture conditions. Digital hemispherical photographs were taken from large corn, soybean and pasture fields and forest plots on 4–5 July 2006 and processed using the CANEYE software to estimate in situ LAI. Estimates derived from PALSAR L-band polarimetric radar backscatter of crop (corn and soybean) fields and forest plots were in good agreement with measured LAI values, but the C-band Advanced Synthetic Aperture Radar imagery showed weak relationships. The study shows that PALSAR L-band polarimetric data have the potential to provide useful estimates of LAI, providing a possible alternative when optical data are limited by cloud cover. However, additional work is required to characterize the temporal variability of the relationship between PALSAR backscatter and LAI over varying soil moisture and soil surface conditions.

A Generalized Radar Backscattering Model Based on Wave Theory for Multilayer Multispecies Vegetation

A generalized radar scattering model based on wave theory is described. The model predicts polarimetric radar backscattering coefficients for structurally complex vegetation comprised of multiple species and layers. Compared to conventional two-layer crown-trunk models, modeling of actual forests has been improved substantially, allowing better understanding of microwave interaction with vegetation. The model generalizes an existing single-species discrete scatterer model and, by including scattering and propagation effects through judiciously defined vegetation layers, enables its application to an arbitrary number of species types. The scatterers within each layer are modeled as finite cylinders or disks having arbitrary size, density, and orientation, as in the predecessor model. The distorted Born approximation is used to represent the propagation through each layer, while scattering from each is modeled as a linear superposition of scattering from its respective random collection of scatterers. Interactions of waves within and between each layer and direct scattering from the ground are accounted for. Validation of the model is presented based on its application to 23 wooded savanna sites located in Queensland, Australia, and comparison with Advanced Land Observing Satellite (ALOS) Phased Arrayed L-band Synthetic Aperture Radar (PALSAR) and National Aeronautics and Space Administration (NASA) Jet Propulsion Laboratory (JPL) Airborne Synthetic Aperture Radar (AIRSAR) data. Results indicate good agreement between simulated and actual backscattering coefficients, particularly at HH and VV polarizations. More discrepancies are found at HV polarizations and can be explained by uncertainties in the knowledge of input parameters, such as inaccuracies in the surface model, surface roughness parameterization, and soil moisture.

Unsupervised Full-Polarimetric SAR Data Segmentation as a Tool for Classification of Agricultural Areas

Versatile, robust and computational efficient methods for radar image segmentation, which preserve the full polarimetric information content, are of importance as research tools, as well as for practical applications in land surface monitoring. The method introduced here consists of several steps. The first step is a (reversible) transform of the full polarimetric radar information content into nine backscatter intensity values. The next steps relate to unsupervised clustering encompassing a simple region-growing segmentation (incomplete and over-segmented) followed by model-based agglomerative clustering and expectation-maximization on the pixels of these segments. Classification is achieved by Markov random field filtering on the original data. The result is a series of segmented maps, which differ in the number of (unsupervised) classes. For a (compatible) supervised approach, only the first and last step have to be applied. Results are discussed for the agricultural areas Flevoland in The Netherlands (AirSAR data) and DEMMIN in Germany, using the NASA/JPL AirSAR system and the DLR ESAR system, respectively. The applications include the use of groundtruth for legend development, the check for groundtruth completeness, and the construction of a bottom-up hierarchy of the characteristics that can be distinguished in the radar data. The latter gives important insights in physics of polarimetric radar backscattering mechanisms. Moreover, the relative importance of crop differences, (full-polarimetric) incidence angle effects and sub-classes (related to factors such as crop varieties, row direction or development stage) may be assessed. The overall classification results range between 84.3% and 98.0%, depending on number of observations dates and radar band(s) used, with higher values for the supervised approach, and substantially more thematic detail for the unsupervised approach.

Polarimetric SAR characterization of man-made structures in urban areas using normalized circular-pol correlation coefficients

Polarimetric Synthetic Aperture Radar (SAR) backscatter from man-made structures in urban areas is quite different than backscatter from predominantly natural areas. Backscatter from natural areas is often reflection symmetric; i.e., characterized by near zero values for covariance matrix off-diagonal terms of the form 〈 S HVS HH ⁎〉, 〈 S HVS VV ⁎〉 and their conjugates. A new approach is proposed to detect scattering from non-reflection symmetric structures using circular-pol, RR-LL, correlation coefficients, | ρ|. This method creates a normalization term, | ρ 0 |, and then forms a ratio, | ρ|/| ρ 0|. The normalization term, | ρ 0|, contains the same diagonal terms of the covariance matrix. The 〈 S HVS HH ⁎〉 and 〈 S HVS VV ⁎〉 off-diagonal terms and their conjugates are purposely set to zero. The ratio, | ρ|/| ρ 0|, is rewritten as a product of separable helicity (τ) and orientation angle ( θ) dependencies. The mathematical form of the τ dependence is a resonant singularity, or pole, term. This pole significantly enhances returns from man-made, high helicity, non-reflection symmetric structures. These structures have values of τ near the resonance value at τ = ± 1. Natural scatterers possess very strong RR / LL symmetry (τ ≈ 0) and the pole response for them is correspondingly weak. The dependence of | ρ|/| ρ 0| on the orientation angle ( θ) is known from previous studies to be useful for measuring urban building alignments (relative to the azimuth direction) and measuring surface topography. The ratio | ρ|/| ρ 0| reduces much of the un-needed image detail of backscatter variations from natural areas of different surface roughness. This image simplification further facilitates detection of localized man-made targets.

Mapping ocean surface features using biogenic slick-fields and SAR polarimetric decomposition techniques

Polarimetric synthetic aperture radar (POLSAR) backscatter from the ocean contains a great deal of information about the physical surface itself. The authors focus on developing a method using polarimetric information and biogenic slick-fields to map selected ocean surface features. The spiral eddy surface features and slick-fields used in this study are those commonly found in waters of the Southern California Bight. To acquire a new dataset, NASA Jet Propulsion Laboratory (JPL) and the Naval Research Laboratory (NRL) jointly flew AIRSAR missions during April 2003 over waters of the Santa Monica Basin and the San Pedro Channel. Conventional single polarisation radar data indiscriminately image the slick-fields, in addition to all unwanted contributions, e.g. intensity modulations owing to the background wave-field and local air–sea interactions. Establishment of an intensity-based threshold to segment the slicks was found to be difficult; particularly under changing wind/wave conditions. In this study several feature and slick mapping algorithms are evaluated. Parameters of the entropy, anisotropy, alpha (〈H/A/α〉) Cloude–Pottier decomposition are used to determine the polarimetric scattering properties of both the biogenic slicks and the ambient ‘clean’ surface. Use of the Cloude–Pottier parameters allows segmentation of shorter Bragg wave effects from longer wave slope modulations. An algorithm combining the Cloude–Pottier decomposition parameters 〈H/A/α〉 with the complex Wishart classifier was evaluated and subsequently selected as the algorithm of choice when using slicks for the production of ocean feature maps. Spiral eddy ocean features, generated by current flow around the southern Channel Islands, were detected and mapped using biogenic slicks as markers for the presence and shape of the eddies.