The European Space Agency's (ESA) BIOMASS mission is a pioneering Earth observation satellite mission launched on April 29, 2025. Utilizing a P-band synthetic aperture radar (SAR), the objective of BIOMASS is to deliver estimates of above-ground forest biomass, forest height (FH), and forest disturbance (FD), with unprecedented accuracy. The mission's primary scientific goal is to quantify the distribution and changes in forest biomass, thereby reducing uncertainties in carbon flux estimates and informing climate models. The satellite's advanced instrumentation and innovative approach allow it to penetrate dense forest canopies, capturing data even in challenging environments. The mission will operate in two distinct phases: the tomographic phase and the interferometric phase, which will support polarimetric interferometric SAR (Pol-InSAR) and tomographic SAR (TomoSAR) processing. Additionally, BIOMASS will provide valuable observational data for ice sheets, deserts, the ionosphere, below canopy topography, and other domains.

ESA BIOMASS will be the first spaceborne P-band SAR, and is designed to produce annual, near-global maps of forest biomass. Biomass-independent scene properties can be mitigated by interferometric pre-processing of the acquisitions, after which backscatter is inverted by a model to estimate biomass, relying on some external reference biomass. The model is simple enough to be invertible, while still representing the spatiotemporal variability of parameters in a global scenario. However, it requires a reliable, global reference biomass dataset, with enough coverage to reflect the variability of the chosen model. The global algorithm is also computationally efficient, so can be deployed in the ground segment. In this paper we briefly discuss the design of the global estimation algorithm for BIOMASS and propose a reference biomass dataset for the algorithm. Then, we discuss the robustness of the proposed algorithm using some reference biomass data for three different parameter variability setups.

The European Space Agency's 7th Earth Explorer mission, BIOMASS, was proposed in 2005 and since then there have been major changes in the scientific and political conditions within which it was conceived, and also within the technology and methodology both of the mission itself and of the complementary systems that BIOMASS will work with. This paper describes some of the most important recent developments in this overall environment of the mission, and how they affect the likely use of data from BIOMASS mission after its launch in 2023 and over its nominal five-year lifetime.

Scheduled for launch in 2023, ESA's seventh Earth Explorer Mission, BIOMASS, will carry the first P-band synthetic aperture radar (SAR) to be flown in space, to gather fully polarimetric acquisitions over forested areas worldwide in interferometric and tomographic modes. This paper presents the algorithms developed to estimate biophysical parameters from BIOMASS measurements and their implementation in the BIOMASS level 2 (L2) prototype processor. The L2 processor will generate global maps of forest Above Ground Biomass (AGB), Forest Height (FH), Forest disturbance (FD). Accurate generation of these products requires the L2 processor to be closely inter-linked with the BIOMASS interferometric processor, in order to produce phase-calibrated interferometric stacks, retrieve sub-canopy terrain topography, and generate a 3D representation of forest structure by use of SAR tomography. AGB estimation results are here shown using BIOMASS-like acquisitions derived from campaign data acquired over six tropical forests in South America and Equatorial Africa.

This paper describes the rationale and development of the estimation techniques for the level-2 data products of the European Space Agency's 7th Earth Explorer BIOMASS mission. BIOMASS is planned for launch in 2023 and will carry the first-ever P-band synthetic aperture radar (SAR) onboard a satellite. It has been designed to produce consistent global maps of the Earth's forests during a nominal five-year lifetime. Fully polarimetric SAR data will be collected and the satellite orbit will be selected for repeat-pass interferometry and tomography in separated mission phases. Mission requirements call for three level-2 data products: above-ground biomass, forest height and forest disturbance. The paper also discusses the expected limitations of the estimation techniques and remaining problems to be addressed.

A new processing technique, i.e., ground cancellation, which removes the ground signal from a pair of interferometric synthetic aperture radar (SAR) images, is used to emphasize the response from above-ground targets. This technique is of particular interest when studying forest canopies using low-frequency signals able to reach the underlying ground, in which case the portion of the signal coming from the ground interferes with the recovery of information about the vegetation. We demonstrate that the power in ground-canceled P-band HV SAR data gives significantly higher correlations with above-ground biomass (AGB) than the interferometric images considered separately. In addition, a significant increase in the sensitivity of backscatter to AGB is observed. Ground-canceled power may then be modeled or regressed to estimate AGB; these possibilities are not discussed here as they will be the topic of forthcoming publications. The effectiveness of this technique is proven through simulations and analysis of real data gathered on tropical forests. The stability of the technique is analyzed under the digital terrain model and baseline control errors, and compensation strategies for these errors are presented.

The ESA BIOMASS mission will be the 7th Earth Explorer measuring the above-ground biomass (AGB) in the world's forests. The current ESA Level-2 (L2) implementation study focuses on defining and implementing the main algorithms for forest parameter retrieval from BIOMASS data. After the first year of L2 study innovative results were achieved: the development of ground cancellation, in particular, has proved to be huge value, since it removes from the data the effects of environmental variability and contributions unrelated to the forest carried in the ground scattering. In this paper the current processor implementation and validation activities of the L2 team will be described.

The structure of forests, in terms of mass and the three-dimensional arrangement of individual trees, is a direct indicator of how much carbon is stored in the ecosystem, which in turn, has a profound effect on how the ecosystem functions and cycles carbon, water, and nutrients (Shugart et al., 2015). There is an increased need to understand local to global storage and dynamics of carbon in terrestrial vegetation, as carbon storage is a prerequisite to understanding the coupling of the biosphere to other components of Earth systems including the climate. The BIOMASS mission of the European Space Agency (ESA), to be launched in 2021-2022 will provide, for the first time, synthetic aperture radar measurements at P-band frequency (~70 cm wavelength) and tomographical (TomoSAR) imaging capability to quantify forest structure and above ground biomass.

The ESA BIOMASS mission will be the first spaceborne mission specifically dedicated to the study of forests. In this framework, ESA is running the “Level-2 implementation study”, focused on defining and implementing the main algorithms for forest parameters retrieval from BIOMASS data. During the first year, the science group involved in the study has carried out R&D activities in order to identify the most viable solutions. In this paper, an overview of the concepts and techniques behind the development of the BIOMASS L2-prototype processor are described, with a focus on each of the main mission products: biomass, forest height and forest disturbance map.

The effectiveness of SAR tomography in estimating forest Above Ground Biomass (AGB) has been repeatedly demonstrated in the recent years. For tropical rain-forests, analysis from the Paracou test site reveals that the best results are achieved when the backscattered power coming from 30m above the ground is considered. As suggested in previous papers, the most likely reason is that ground scattering acts as a disturbing factor for forest biomass retrieval, as it depends on a number of parameters (like topography, moisture), that do not relate to forest biomass. In this paper we further test this hypothesis by proposing the concept of interferometric ground notching. By taking the difference between two phase calibrated, ground-steered, SAR SLC images a third image is obtained where ground scattering contributions are canceled out, hence the name ground-notched SLC. Results indicate that ground-notched data can effectively retain the features of vegetation-only scattering, including its polarimetric signature and correlation with AGB.

The BIOMASS mission gives completely new challenges in external calibration arising from the orbital pattern needed for the tomographic and Pol-InSAR phases of the mission, the strong effects of the ionosphere at P-band, and the lack of pre-existing P-band data except over very limited parts of the globe. Together these create problems that can only be solved by combining infrequent visits to instrumented calibration sites with systematic exploitation of the properties of distributed targets and targets of opportunity. Proposed approaches to performing radiometric and polarimetric calibration are described, together with meeting geolocation accuracy requirements.

This paper presents a new algorithm for forest biomass estimation from P-band synthetic-aperture radar (SAR) backscatter data, notch-filtered at ground-level. A semi-empirical model is fitted to spatial and polarization trends in the backscatter data and no reference biomass data are needed for training. An evaluation on airborne P-band SAR data from a tropical test site in Gabon results in a root-mean-square error lower than 20% and a correlation better than 90%.

The Earth Explorer Biomass mission will provide the scientific community with accurate maps of tropical, temperate and boreal forest biomass, including height and disturbance patterns. This information is urgently needed to improve our understanding of the global carbon cycle and to reduce uncertainties in the calculation of carbon stocks and fluxes associated to the terrestrial biosphere. It is also crucial for approaches to managing climate, such as the UNFCCC initiative known as Reducing Emissions through Degradation and Deforestation (REDD+), aimed at climate change mitigation through conservation and better management of tropical forests The required measurements are forest biomass and forest height at resolution of 200 m, and detection of deforestation at 50 m. Global maps of biomass are required with accuracy of 20% (or l0 t ha-1when above-ground biomass are less than 50 t ha-1). To achieve this Biomass will be implemented as a P-band SAR mission. It will exploit the unique sensitivity of P-band SAR together with advanced retrieval methods including polarimetric interferometry (Pol-InSAR) and SAR tomography to measure biomass, height and disturbances across the entire biomass range every 6 months. The mission will also support important secondary objectives, including sub-surface imaging in arid zones, production of a bare-earth DTM and ice applications.

The vertical electron density distribution can have a significant impact on ionospheric measurements of Total Electron Content (TEC) derived from estimates of Faraday rotation (FR) with spaceborne low-frequency SAR data. Using vertical electron density data and the IGRF-12 magnetic field model, simulations of FR under different vertical electron density distributions for an L-band SAR have been performed. The usual approximation that the geometric and magnetic field parameters can be estimated at 400 km when converting FR to TEC is shown to cause significant errors, especially for regions at low latitude and near the zero FR line.

A significant amount of the data acquired by sun-synchronous space-borne low-frequency synthetic aperture radars (SARs) through the postsunset equatorial sector are distorted by the ionospheric scintillations due to the presence of plasma irregularities and their zonal and vertical drift. In the focused SAR images, the distortions due to the postsunset equatorial ionospheric scintillations appear in the form of amplitude and/or phase “stripe” patterns of high spatial frequency aligned to the projection of the geomagnetic field onto the SAR image plane. In this paper, a methodology to estimate the height and the drift velocity of the scintillations from the “stripe” patterns detected in the SAR images is proposed. The analysis is based on the fact that the zonal and vertical drift of the plasma irregularities are, at the equatorial zone, perpendicular to the geomagnetic field which is almost parallel aligned to the orbit. The methodology takes advantage of the time lapse and change of imaging geometry across azimuth subapertures. The obtained height estimates agree well with the reference measurements and independent estimates reported in the literature, while the drift velocities appear slightly overestimated. This can be attributed to a suboptimum geometry configuration but also to a decoupling of the ambient ionosphere and the plasma irregularities.

This is a selection of results of the North State project, that demonstrate how innovative methods applied to the new Sentinel data streams can be combined with models to monitor carbon and water fluxes for pan-boreal Europe.

The P-band synthetic aperture radar of the European Space Agency BIOMASS mission will be affected by ionospheric phase scintillation at high latitudes, which introduces a random high-order azimuth phase error. The dependence of the performance of the phase gradient autofocus (PGA) algorithm for scintillation compensation on the strength of ionospheric turbulence and the signal-to-clutter ratio (SCR) is analyzed. In order to keep resolution degradation below 2%, the SCR must exceed 16 and 20 dB for turbulence strengths CkL = 1033and 1034, respectively. For large values of CkL, phase scintillation adds significantly to post-PGA degradation in the integrated and peak sidelobe ratios. Simulations based on scenes derived from PALSAR data demonstrate the effectiveness of PGA.

A correction methodology for distortions induced by ionospheric scintillation on fully polarimetric synthetic aperture radar (SAR) data is proposed. The correction is based on deriving the phase distortion induced by the ionosphere from Faraday rotation estimates. The estimated phase distortion is then used for correction. In order to compensate the phase and time-Doppler history distortions, the correction has to be performed at the slant range of the ionospheric layer, i.e., on partially focused single-look complex data. Accordingly, the performance of the proposed correction methodology depends, among other factors, on knowledge of the altitude of the effective ionospheric layer (assuming the thin ionospheric layer model). Its estimation from the SAR data itself is therefore also addressed. The methodology was applied and validated on simulated P-band data for various ionospheric conditions and on real L-band data acquired by the Advanced Land Observation Satellite Phased Array L-band SAR (PALSAR).

Radio waves traversing the Earth's ionosphere suffer from Faraday rotation with noticeable effects on measurements from lower frequency space-based radars, but these effects can be easily corrected given estimates of the Faraday rotation angle, i.e., Ω. Several methods to derive Ω from polarimetric measurements are known, but they are affected by system distortions (crosstalk and channel imbalance) and noise. A first-order analysis for the most robust Faraday rotation estimator leads to a differentiable expression for the bias in the estimate of Ω in terms of the amplitudes and phases of the distortion terms and the covariance properties of the target. The analysis applies equally to L-band and P-band. We derive conditions on the amplitudes and phases of the distortion terms that yield the maximum bias and a compact expression for its value for the important case where Ω = 0. Exact simulations confirm the accuracy of the first-order analysis and verify its predictions. Conditions on the distortion amplitudes that yield a given maximum bias are derived numerically, and the maximum bias is shown to be insensitive to the amplitude of the channel imbalance terms. These results are important not just for correcting polarimetric data but also for assessing the accuracy of the estimates of the total electron content derived from Faraday rotation.

For long-wavelength space-based radars, such as the P-band radar on the recently selected European Space Agency BIOMASS mission, system distortions (crosstalk and channel imbalance), Faraday rotation, and system noise all combine to degrade the measurements. A first-order analysis of these effects on the measurements of the polarimetric scattering matrix is used to derive differentiable expressions for the errors in the polarimetric backscattering coefficients in the presence of Faraday rotation. Both the amplitudes and phases of the distortion terms are shown to be important in determining the errors and their maximum values. Exact simulations confirm the accuracy and predictions of the first-order analysis. Under an assumed power-law relation between σhvand the biomass, the system distortions and noise are converted into biomass estimation errors, and it is shown that the magnitude of the deviation of the channel imbalance from unity must be 4-5 dB less than the crosstalk, or it will dominate the error in the biomass. For uncalibrated data and midrange values of biomass, the crosstalk must be less than -24 dB if the maximum possible error in the biomass is to be within 20% of its true value. A less stringent condition applies if the amplitudes and phases of the distortion terms are considered random since errors near the maximum possible are very unlikely. For lower values of the biomass, the noise becomes increasingly important because the σhvsignal-to-noise ratio is smaller.

Spaceborne linearly polarimetric synthetic aperture radar (SAR) provides information about surface properties and scatterer types by measuring the covariances between images formed using orthogonally polarized radio waves. However, ionospheric Faraday rotation (FR) and system calibration errors alter the balance in the polarimetric channels and distort the covariance matrix. Since FR angles are greater at lower radio frequencies, this paper focuses on the Biomass satellite, which is the European Space Agency's 7th Earth Explorer mission and will carry a P-band (435 MHz) polarimetric SAR. Its primary objective is to measure forest biomass density by combining estimates derived from the polarimetric covariance matrix with estimates based on measuring forest height with polarimetric interferometry and exploiting height-biomass allometric relations. The accuracy of four methods for estimating FR from polarimetric SAR is assessed using simulated images of forest with a range of biomass densities and system errors (H/V channel imbalances, antenna crosstalk, and noise) and over a large spread of FR angles. All methods have biases dependent on the FR, the relative phases of the crosstalk components, and the channel imbalance phase. The best performing method estimates FR to better than 5°under worst case system errors at all FR angles, so the accuracy of biomass density estimates should not be significantly affected. However, crosstalk phases are predicted to vary greatly across the Biomass antenna beam, and this could potentially cause the FR bias to vary by several degrees across the swath. These effects may occur even for the small values of FR met in L-band data.

The European Space Agency is conducting studies for a low-earth orbiting polarimetric synthetic aperture radar called BIOMASS to provide global measurements of forest biomass and tree height. Phase scintillation across the synthetic aperture caused by ionospheric irregularities can degrade the impulse response function (IRF) and cause squinting, and its temporal variation can cause decorrelation in repeat-pass interferometry. These effects are simulated for a range of conditions for the baseline BIOMASS system configuration using the Wideband model of scintillation, which predicts that for a dawn–dusk orbit, impacts of scintillation over forest regions are negligible under all conditions except at high latitudes in the North American sector under high sunspot activity. In this sector, single-look IRFs have mean integrated sidelobe ratios (ISLRs) and peak sidelobe ratios (PSLRs) better than 0 and${-}{\rm 5}~{\rm dB}$, respectively, at 90% confidence interval under median solar activity up to the northern tree line (${\sim}{\rm 70}^{\circ}$geomagnetic). Degradation in the mean 3-dB resolution of up to 10% is predicted, with mean absolute azimuth shifts of the IRF peak of up to 2 m, which increases to 5 m at high sunspot number. Similar values are found for the dawn and dusk sides, and seasonal variations are negligible for latitudes below the tree line. Repeat-pass interferometric image pairs maintain coherence${>}{\rm 0.8}$up to 50$^{\circ}~{\rm N}$under median sunspot conditions. Four-look processing improves the ISLR and PSLR by several decibels, but causes significant degradation of the 3-dB resolution due to incoherent averaging of images with different random azimuth shifts.

A correction method for distortions induced by ionospheric scintillation effects on P-band quad-pol synthetic aperture radar (SAR) data acquired in polar and high latitude regions is presented. For this the estimation of the Faraday rotation (FR) is converted to ionospheric phase and used for the correction of the scintillation. The correction is performed on partially focused SLC data in order to compensate the ionosphere-induced phase and the distortion of the time-Doppler history. The degree of defocusing depends on the altitude of the ionosphere. The performance of the new correction method is tested on P-band simulated data for various ionospheric scenarios.

This paper provides an overview of the BIOMASS Mission End-to-End simulator (BEES) and of the mission performance analysis performed with it. The end-to-end performance, in terms of biomass estimates error, is close to the 20% error goal set for the mission. The main sources of errors are temporal decorrelation and the limited available bandwidth, while system induced errors have a negligible impact on the final performance.

The BIOMASS mission is designed to provide unique information on the biomass in the world's forests at spatial and temporal resolutions suitable for characterizing their dynamics and their contribution to carbon cycle estimates. To achieve this it combines biomass estimates from direct inversion of polarimetric backscattering coefficients with Pol-InSAR forest height estimates. The mission will also support important secondary objectives, including sub-surface imaging in arid zones, production of a bare-earth DTM and ice applications, and is optimized to be robust against environmental and ionospheric disturbances.

Information on above-ground biomass and its changes is of fundamental importance for carbon cycle and climate calculations. The difficulty of making ground-based inventories of biomass, especially in the tropics, has therefore led to serious attempts to estimate biomass from space. Active technologies hold the most promise. Lidar maps of forest height derived from IceSAT data have underpinned two recent pan-tropical biomass maps, using indirect statistical regression methods. More direct measurements of biomass are possible using radar, and can be combined with height measurements by a mission that supports a Pol-InSAR mode. Physics and experiment both indicate that the longest wavelength possible should be used, which is P-band. The BIOMASS mission is driven by these considerations, and is uniquely fitted to measure biomass, but will gain significantly from combination with both spaceborne L-band data, expected to be available during the mission lifetime, and with airborne lidar measurements.

This paper quantifies the probable impacts of ionospheric scintillation on the proposed European Space Agency BIOMASS synthetic aperture radar (SAR). BIOMASS is a 435 MHz quad-polarised.SAR in a dawn-dusk orbit, with a primary objective of measuring woody biomass density and tree heights with near-global coverage. The climatological model WBMOD is used to generate multiple random ionospheric phase screens, from which are derived SAR impulse response functions (IRFs) for a range of locations, times and solar and geomagnetic activity levels. The statistics of IRF degradation are presented in terms of the integrated and peak sidelobe ratios, resolution increases, azimuthal shifts and reductions of the main lobe peak. The ionospheric impacts for global forest regions are found to be negligible under all conditions except at high latitudes in the North American sector under high solar and geomagnetic activity. Consequences for ice monitoring (at higher latitudes) will be much more widespread.

The BIOMASS mission is designed to map the full range of the world's above-ground forest biomass, for the needs of national scale inventory and global carbon flux calculations. This objective is achieved with advanced P-band SAR techniques. The P-band biomass measurement concept was based on previous work over the past two decades. During the preparatory phase, new campaigns have been conducted to address critical issues on the biomass retrieval algorithms, over tropical and boreal forests. The collected datasets comprise accurate and complete sets of in situ data and advanced P-band SAR data. This paper presents the retrieval algorithms developed using the collected datasets.

This paper discusses the implementation of an end- to-end simulator for the BIOMASS mission. An overview of the system architecture is provided along with a functional description of the modules that comprise the simulator.

Spaceborne synthetic aperture radar (SAR) systems operating at lower frequencies, such as P-band, are significantly affected by Faraday rotation (FR) effects. A novel algorithm for calibrating the circular-transmit-and-linear-receive (CTLR) mode spaceborne compact polarimetric SAR using mixed calibrators is proposed, which is able to correct precisely both FR and radar system errors (i.e., channel imbalance and crosstalk). Six sets of mixed calibrators, consisting of both passive calibrators and polarimetric active radar calibrators (PARCs), are investigated. Theoretical analysis and simulations demonstrate that the optimal calibration scheme combines four polarimetric selective mixed calibrators, including two gridded trihedrals and two PARCs, together with total-electron-content measurements by the Global Navigation Satellite System system.

The European Space Agency (ESA) released a Call for Proposals for the next Earth Explorer Core Mission in March 2005, with the aim to select the 7thEarth Explorer (EE-7) mission for launch in the next decade. Twenty-four proposals were received and subject to scientific and technical assessment. Six candidate missions were selected and further investigated in the preliminary feasibility studies (Phase 0). One of these missions is BIOMASS, which has recently been selected to proceed to Phase-A. BIOMASS is a response to the urgent need for greatly improved mapping of global biomass and the lack of any current space systems capable of addressing this need.

Spaceborne polarimetric synthetic aperture radar systems operating at lower frequencies, such as P-band, are significantly affected by Faraday rotation (FR). A new set of FR estimators is derived from the off-diagonal terms in the measured covariance matrix of a distributed target. These estimators have a phase ambiguity of period , instead of /2 as for the published estimators, and this ambiguity can be completely resolved for arbitrarily large values of FR using total electron content maps derived from Global Navigation Satellite System measurements. Simulations show that one of the new estimators has particularly high resistance to system noise and channel amplitude imbalance but greater sensitivity to channel phase imbalance than the published estimators. Hence, the expected values of residual system distortion after calibration may affect the choice of estimator.

Faraday rotation and scintillation will seriously affect P-band SAR images if methods are not used to deal with them. There are published methods for correcting Faraday rotation when polarimetric data are available, but less is known about scintillation. Scintillation is unlikely to present a problem if data acquisitions at low and mid latitudes avoid the post-sunset equatorial region, but correction procedures are likely to be needed in the auroral zones. Such correction is possible using autofocus methods; this is demonstrated using the Phase Gradient Algorithm applied to SAR imagery seriously affected by simulated scintillation.

In this paper the feasibility of using passive microwave data from the Special Sensor Microwave/Imager (SSM/I) satellite sensor to retrieve information on the snow water equivalent (SWE), representing the mass per unit area of water contained within a snowpack, is investigated for an area of central Siberia. A modelling chain combining a dynamic model of a snowpack and a model of the brightness temperature obtained from a passive microwave sensor such as SSM/I was used to generate predictions of the difference in brightness temperature between 19 and 37 GHz for mid-Siberia, and these results compared to the actual SSM/I data.

The variability of C-band one-day tandem coherence measurements of forest is analyzed both statistically and with the aid of a soil-vegetation-atmosphere transfer model coupled to models for dielectric, backscattering coefficient, and coherence. Coherence depends strongly on both moisture and wind conditions. The dependence between coherence and moisture fluctuations in the soil and canopy is simulated using a simplified coherence model neglecting canopy coherence; trends are similar to those observed in measured coherence, but the range of variability observed in the satellite data is underestimated. Improved consistency with the data is achieved by using empirical estimates of canopy coherence derived from mature forest, but still with underestimated variability. The canopy coherence values cannot be predicted from the environmental model, and anomalous observations prevent the formulation of a reliable empirical predictor based on weather data; this could potentially be improved if higher time resolution wind data were made available. The study also suggests that canopy dielectric can vary rapidly in daylight, and that current models for dielectric may be inadequate to capture the effect of this behavior on coherence.

We present results from experiments carried out in the ground-based synthetic aperture radar (GB-SAR) facility at the University of Sheffield to ascertain the role of polarimetric interferometry in crop height retrieval. To this end, a mature wheat canopy, grown in outdoor conditions, was reassembled inside the GB-SAR chamber and imaged at C-band using a two-dimensional scan. This allowed fully polarimetric tomography and interferometry. Interferometry using the VV, HH, and VH polarization states shows that the HH and VH interferograms retrieve a height close to the top of the soil layer for all angles of incidence considered, whereas the height retrieved from the VV interferogram increases with angle of incidence. The use of a Pauli basis gives poor results, due to the different location of the scattering phase centers in the VV and HH channels. The use of arbitrary polarization states shows that the top of the soil can be very accurately estimated using left-circular polarization, whereas, for angles of incidence close to 45/spl deg/, a polarization state similar to VV can be used to retrieve the top of the canopy; hence crop height can be recovered as the difference of these two interferometric heights. Polarimetric coherence optimization techniques are also studied. Unconstrained coherence optimization gives very unstable results, due to the small number of available samples. Constrained optimization results in stable retrieved heights, and the retrieved polarization states agree well with the polarization synthesis results.

Ground based facilities allow for tightly controlled experiments which are very useful to identify the suitability and limitations, as well as the development of observational techniques. In this work, we report on the use of an indoor anechoic chamber, part of the GB-SAR facility, for polarimetric interferometry studies of vegetation. We give a brief description of the system, pointing out a few differences with common airborne I and spaceborne systems, present results of single polarisation interferometry using the H/V basis. These results show that for large angles of incidence, the height difference between the VV-VV and HH-HH interferograms can be used as an estimate of crop height. The use of coherence optimisation (both constrained and unconstrained) is also investigated. These techniques find polarisation states that allow for the accurate estimation of crop height. Some concluding remarks on the issues and prospects of ground based systems for this kind of work are also presented.

This paper describes an algorithm aimed at monitoring the soil moisture and the growth cycle of wheat fields using radar data. The algorithm is based on neural networks trained by model simulations and multitemporal ground data measured on fields taken as a reference. The backscatter of wheat canopies is modeled by a discrete approach, based on the radiative transfer theory and including multiple scattering effects. European Remote Sensing satellite synthetic aperture radar signatures and detailed ground truth, collected over wheat fields at the Great Driffield (U.K.) site, are used to test the model and train the networks. Multitemporal, multifrequency data collected by the Radiometer-Scatterometer (RASAM) instrument at the Central Plain site are used to test the retrieval algorithm.

Radiative transfer models have been widely used to interpret the radar backscatter from forested areas. Most of these models are based on an iterative solution of the radiative transfer equation, usually solved up to first or second order, thus taking into account single and double scattering. Although this method leads to results agreeing well with copolarized backscatter measurements, it produces less accurate estimates for horizontal-vertical (HV) polarization. This paper presents a radiative transfer backscatter model that accounts for multiple scattering by using the discrete ordinate and eigenvalue method applied to a layered medium. Using parameters derived from an architectural tree model, calculations at C- and L-band are compared with HV data acquired for a maritime pine forest in the southwest of France during the Spaceborne Imaging Radar-C missions. Good agreement is found at C-band for all values of forest biomass, and reasonable agreement at L-band for high biomass, when the soil backscatter plays a minor role. For low biomass, the L-band modeling is inadequate because of difficulties in estimating the soil backscatter. Comparison with calculations from a first-order radiative transfer model shows that multiple scattering is significant, especially at C-band.

This paper describes an algorithm aimed at monitoring the soil moisture and the growth cycle of wheat fields using radar data. The algorithm is based on neural networks trained by an electromagnetic model and multitemporal ground data measured on fields taken as a reference. The retrieval procedure is tested using mutitemporal signatures collected at a test site.

The current understanding of the part that the boreal region plays in carbon flux and ultimately in climate change is poor, partly because of the lack of information on disturbance (fires and logging), freeze-thaw transitions (especially with respect to permafrost) and overall vegetation distribution, and party because all of these issues have never been addressed before in an integrated manner. The main objectives of the EC-project SIBERIA-II are the integration and combination of multi-sensor remotely sensed data and ecological regional models in order to assess the impact of terrestrial biota on the budget of major greenhouse gases (GHGs) in Northern Eurasia and the demonstration of viability of full carbon accounting (including CO/sub 2/, CO, CH/sub 4/, N/sub 2/O, NO/sub x/) using Dynamic Vegetation Models (DVMs) and multi-sensor Earth Observation (EO) instruments.

Polarimetric X- and C-band measurements by the University of Sheffield ground-based synthetic aperture radar (GB-SAR) indoor system provide three-dimensional images of the scattering processes in wheat canopies, at resolutions of around a wavelength (3-6 cm). The scattering shows a pronounced layered structure, with strong returns from the soil and the flag leaves, and in some cases a second leaf layer. Differential attenuation at horizontal (H) and vertical (V) polarization, due to the predominantly vertical structure of the wheat stems, gives rise to marked effects. At both C and X bands, direct return from the canopy exceeds the soil return at large incidence angles for VV polarization, but is comparable to or less than the soil return in all other cases. At HV, the apparent ground return is probably due to a double-bounce mechanism, and volume scattering is never the dominant term. Direct sensing of the crop canopy is most effective at X band, VV, and large incidence angles, under which conditions the return is dominated by the flag leaf layer. Field measurements with the outdoor GB-SAR system suggest, however, that for sensitivity to biomass and reduced susceptibility to disturbances by rainfall, a two-channel C-band system operating at a medium range of incidence angles is preferred.

Several proposed near-future spaceborne radar missions, such as the Advanced Land Observing Satellite (ALOS) and TerraSAR, will include an L-band instrument. At such low frequencies, the Faraday rotation in the ionosphere, which rotates the polarization plane of the radar signal, becomes an important consideration in instrument design. In this paper, both simple analytic approximations and numerical models are used to derive likely values of Faraday rotation and determine their impact on polarimetric imagery and derived products. One-way rotations exceeding 5/spl deg/ are likely to significantly reduce the accuracy of geophysical parameter recovery, such as forest biomass. On average, Faraday rotation can be neglected at solar minimum, but correction methods are needed at other times of the solar cycle and under disturbed conditions. Methods for implementing such corrections based on estimates of the Faraday rotation and prerotation of the transmitted signal are described.

An algorithm, based on an electromagnetic model and a neural network, aimed at monitoring the multitemporal evolution of wheat fields, is described. Three different sites are used to validate the model, provide reference ground data, and test the algorithm.

An explicit form of the linear multichannel synthetic aperture radar (SAR) intensity filter, which preserves radiometry while optimally reducing speckle is derived, together with a compact expression for the theoretical gain in equivalent numbers of looks (ENLs). The filter can be applied to mixed data types, which is demonstrated using a combination of ERS and JERS satellite data, and confirms the filter performance predicted by the theory. Tests indicate that a simplified form of the filter, which neglects correlation between images, gives an ENL only slightly less than optimal, while being much easier to implement. Exact analysis of the effect of estimating filter weights shows that the linear increase in ENL with the number of images predicted for the ideal filter does not occur. In practice, the ENL is affected by the window size used to estimate the weights and saturates as the number of images increases. An efficient recursive form of the filter is described, which is most naturally applied to multitemporal data for the practically important case where the current image is uncorrelated with previous images in a data sequence.

Using ERS scatterometer data effectively to relate large scale surface processes to driving parameters such as vegetation type, weather, soil, etc., requires the data be referenced to a regular geographic grid. The averaging involved may conceal systematic variation in /spl sigma//sup 0/ in time, space, and across incidence angle. The problems in identifying the sources of variation at a given position are discussed and the relative magnitudes of the different variability components are evaluated. Variability indices can be defined at each gridpoint, of which the most useful appears to be the coefficient of variation after model-based correction for incidence angle effects. Images of variability indicate that for about 75% of the land surface, the mean /spl sigma//sup 0/ images can be considered representative. However, some cover types exhibit continual, significant variability on short timescales, particularly grasslands and sandy deserts. Other cover types display seasonal variation, which appears to be related to vegetation, snow cover, and the freeze/thaw cycle. Variability measures also provide clear indications of occasional data problems even where no data quality flags are set.

A comprehensive multilayer second-order radiative transfer model, driven entirely by intensive field observations, is used to show that second order terms contribute at most 0.5 dB to the backscattering coefficient at all polarizations from wheat and barley throughout the growing season and 1 dB to the copolar response of oilseed rape. Under these circumstances, an equivalent integrable first-order model, with coefficients derived from full model runs, can be formulated. This allows the role of each of the plant components in attenuating and scattering the radar signal to be clarified, and provides a basis for quantitative comparison of observed ERS-2 backscatter values with model calculations, taking full account of measurement uncertainties. Discrepancies between the two suggest that effective attenuation through mature cereal crops is overestimated by the model. This appears to be either an intrinsic failure of the radiative transfer formulation or (more likely) due to an inadequate adaptation of the notion of crop coverage to the microwave case. Nonplanarity of leaves is an important source of model error for oilseed rape.

Examination of the physical background underlying the ERS response of forest and analysis of time series of ERS data indicates that the greater temporal stability of forest compared with many other types of land cover presents a means of mapping forest area. The processing chain necessary to make such area estimations involves reconstruction of an optimal estimate of the backscattering coefficient at each pixel using temporal and spatial filtering so that classification rules derived from large scale averaging are applicable. The rationale behind the filtering strategy and the level of averaging needed is explained in terms of the observed multitemporal behavior of forest and nonforest areas, much of this analysis is generic and applicable to a wide range of situation in which significant information is carried by multitemporal features of the data. The choice of decision rules is based on the forest observations, with the added requirement for robustness. The performance of a classifier based only on change is assessed on a range of test sites in the UK, Finland, and Poland. Error sources in this classifier are identified, and the possibility of improving performance by including radiometric information in the mapping strategy is discussed. Brief discussions of how the classification is affected by the addition of coherence and how the processing chain would need to be modified for other forms of satellite data are included.

The potential of spaceborne synthetic aperture radar (SAR) for monitoring tropical forest areas is assessed, using three ERS images from the Tapajos region of Amazonia gathered in 1992 and a single JERS-1 image of the same area acquired in 1993. The multitemporal ERS-1 data indicate that primary forest areas display a very stable radar backscattering coefficient (/spl sigma//sup 0/), while in some cases, disturbed areas (nonforest and regenerating forest) exhibit changes that appear to be associated with soil moisture variations. To counteract /spl sigma//sup 0/ distortions caused by topography, change detection based on ratios of intensity images (or differences of log images) provides a more useful discrimination approach than /spl sigma//sup 0/ variations in single images. Change detection techniques are compared, and their ability to classify primary and disturbed forest is quantitatively assessed, assuming that a land cover map inferred from a 1992 Landsat thematic mapper (TM) image is correct. Even in the best case, less than 50% of the disturbed forest region is detected in the ERS-1 images. This figure may be improved by more frequent image acquisition, but there are fundamental limitations in using C-band data since the effects of soil moisture changes on /spl sigma//sup 0/ are masked once even comparatively low levels of standing biomass are present. At the longer wavelength of JERS-1, much better discrimination is possible, but the correction of topographic distortions is likely to present problems.

Methods to evaluate the performance of segmentation algorithms for synthetic aperture radar (SAR) images are developed, based on known properties of coherent speckle and a scene model in which areas of constant backscatter coefficient are separated by abrupt edges. Local and global measures of segmentation homogeneity are derived and applied to the outputs of two segmentation algorithms developed for SAR data, one based on iterative edge detection and segment growing, the other based on global maximum a posteriori (MAP) estimation using simulated annealing. The quantitative statistically based measures appear consistent with visual impressions of the relative quality of the segmentations produced by the two algorithms. On simulated data meeting algorithm assumptions, both algorithms performed well but MAP methods appeared visually and measurably better. On real data, MAP estimation was markedly the better method and retained performance comparable to that on simulated data, while the performance of the other algorithm deteriorated sharply. Improvements in the performance measures will require a more realistic scene model and techniques to recognize oversegmentation.

A unified approach to phase and cross-talk calibration of polarimetric data which can be applied to calibrating scattering matrix data or to extraction of the descriptors of distributed targets is described. It relies on the scene being dominated by targets with uncorrelated like and cross-polarized backscattering coefficients, but provides cross-talk calibration of targets for which this is not true. The algorithm needs unsymmetrized data, but uses only quantities derived from the covariance matrix of large areas. It makes no assumptions about system reciprocity, permits ready interpretation of the terms in the calibration procedure, allows comparison of the relative magnitude of the system-induced mixing of terms in the observed covariance matrix, is noniterative, and produces indicators which allow testing of whether it meets its own underlying assumptions. The linear distortion model is shown to lead to an inconsistent system of equations; this inconsistency can be removed by introducing an extra parameter which has properties expected of system noise. The modulus of the copolarized correlation coefficient, which is important in polarimetric classification and as a phase descriptor, is shown to be invariant under all effects embodied in the linear distortion model. Calibration of the scattering matrix data is based on a minimum least squares principle. This suggests that current methods of symmetrization are not optimal. The same analysis shows that estimates of parameters needed to form an equivalent reciprocal system are also nonoptimal. The method is more general than the well-known van Zyl algorithm for cross-talk removal, and permits an analysis of the conditions under which the van Zyl algorithm will yield valid results. Correction of phase distortion induced by channel imbalance Is treated as an optional extra step relying on a known HH-VV phase difference in some region of the image. Results from the algorithm are discussed using scattering matrix data from the 1989 MAESTRO campaign.>

The feasibility of matching templates representing map features to features in synthetic aperture radar images is investigated. The performance of template matching using cross-correlation (which has no knowledge of speckle) is compared with a correlation measure, based on a Kolmogorov-Smirnov test of fit, which does take speckle into consideration. This is shown to perform better than cross-correlation in simulated images. However, neither method performs well on real data. This is due to the exponential speckle model being inappropriate for linear features in real data. Alternative methods for matching map features to SAR images are suggested.>