The application of polarimetric Synthetic Aperture Radar (SAR) to forest observation for mapping, classification and parameter estimation (especially biomass) has a relatively long history. The radar penetration through forest volume, and hence the multi-layer nature of scattering models, make fully polarimetric data the observation space enabling a robust and full inversion of such models. A critical advance came with the introduction of polarimetric SAR interferometry, where polarimetry provides the parameter diversity, while the interferometric baseline proves a user-defined entropy control as well as spatial separation of scattering components, together with their location in the third dimension (height). Finally, the availability of multiple baselines leads to the full 3-D imaging of forest volumes through TomoSAR, the quality of which is again greatly enhanced by the inclusion of polarimetry. The objective of this Chapter is to review applications of SAR polarimetry, polarimetric interferometry and tomography to forest mapping and classification, height estimation, 3-D structure characterization and biomass estimation. This review includes not only models and algorithms, but it also contains a large number of experimental results in different test sites and forest types, and from airborne and space borne SAR data at different frequencies.

In this paper, we consider the problem of radar estimation of forest canopy height in regions with dense forests and severe topography. We combine a reference digital elevation model with multiple satellite baselines from ascending and descending orbits to develop a merging algorithm relating single pass interferometric coherence to forest canopy height. We first describe the algorithm and processing steps used for height estimation and then apply the technique to a mountainous study site in British Columbia, Canada, using data from the Tandem-X satellite pair. We devise a new masking scheme to isolate potential problem areas in sloped terrain and apply the new merging algorithm by using multiple Tandem-X tracks to overcome the gaps left due to the masking procedure. The radar height products are validated by using a network of ground forest measurement sites and supporting lidar. The regression statistics show an r2 of 0.70 and rmse of 4.1 m between the radar and the field measured heights. By examining height errors, we implement a new test for the presence of canopy extinction, or subcanopy surface scattering, and demonstrate that in the dense and mountainous forests of British Columbia, there are significant canopy extinction effects in X-band imagery.

In this paper, we consider the problem of true transmitter polarization state estimation for circular compact SAR modes. We employ two methods-the first using point targets such as trihedral reflectors, and the second, a new method, based on the Pauli phase observed over distributed targets like forest canopy. We first show how compact modes allow estimation of the Pauli phase for reflection symmetric scatterers. We show that this phase remains remarkably constant over forest canopies, depending primarily on the dielectric constant of constituent volume particles. We show that small imperfections in the transmit polarization state then lead to large errors in this phase estimate. These errors can then be used as the basis for a calibration strategy that allows estimation of a set of candidate true transmitter states. These can then be compared with the trihedral estimates for validation. We illustrate using L-band compact data from the JAXA ALOS-2 satellite and C-band data from the ISRO RISAT-1 satellite.

RADARSAT-2 SAR data was used to develop a monitoring program for Canadian forest lands with the aim to provide information on forest harvesting. Study sites in British Columbia and New Brunswick Canada were selected. RADARSAT-2 MultiLook Fine mode, acquired from mid-June through mid-September, from 2011 to 2015, was analyzed with the aim to detect forest disturbances.Due to large data volumes and the need for efficiency, an automated end-to-end solution was implemented. The automated solution included image coregistration, temporal filtering, detection of forest disturbances, and delineation of the disturbances. To reduce the detection of false positives, a non-forest mask was developed that entailed a combination of CanVec data that delineated areas such as water bodies, roads, and urban/industrial areas and SAR-derived information such as layover and scattering from urban areas.To assess the performance of the change detection algorithm, the RADARSAT-2 changes were compared to tree-loss information from the Canadian Forest Service (CFS). Since CFS information was representative of annual changes, but the RADARSAT-2 derived changes were representative of summer-only changes, there were discrepancies between the RADARSAT-2 data and the CFS data. Notwithstanding these discrepancies, the detection performance, based on the RADARSAT-2 and CFS changes overlapping by at least 50%, was better than 74%, with one exception at 62%.The tree loss area derived from RADARSAT-2 was compared to the CFS data. Other than one case, the RADARSAT-2 area was greater than the CFS area. The larger RADARSAT-2 areas was attributed to the auto-generation of areas-of-change. When the changes were visually digitized from the RADARSAT-2 imagery, the area from RADARSAT-2 was within approximately 8% to 17% of the CFS area.

Forest height is a key measurement for determining aboveground carbon in forests and forest biomass. DLR's TANDEM-X mission provides paired scenes that can be used for forest height measurement. We have demonstrated this in [1,2,3] for scenes with some moderate topography. This paper discusses issues for estimating forest canopy heights in mountainous terrain. Multiple dates and baselines of TanDEM-X data were obtained for a study site with much more severe topography and taller trees in BC, Canada. The results of analysis of large-slope induced errors and mitigation of some of these effects are discussed below.

In this paper we describe remote sensing methods with optical and radar sources to monitor aboveground carbon in forests. We move from forest/non-forest mapping, through forest type mapping (conifer, deciduous and mixed) to major species mapping. Biomass combined with species is required to compute the aboveground carbon. Biomass for forests needs accurate height measurements and allometric equations relating height, species and biomass. LIDAR data and TanDEM-X data were used for height measurement. We close with a discussion of accuracies and sources of error in making carbon maps with remote sensing and allometric models.

RADARSAT-2 SAR data were used to develop a monitoring program for Canadian forest lands with the aim to provide information on forest cutblocks and to develop algorithms for the detection of forest disturbances. Three study sites were selected in Canada and were representative of different forest types and terrain. Due to large data volumes and the need for efficiency, an automated end-to-end solution was implemented. The automated solution included image coregistration, temporal filtering, detection of forest disturbances, and delineation of the disturbances. A critical component of the overall solution was the validation: cutblock information and Landsat imagery were used. Overall there was good agreement between the changes detected in the RADARSAT-2 data and the validation data. False positives were detected that were mainly due to changes from non-forestry natural sources and anthropogenic sources. In general, the separation of the forest and the non-forest areas-of-changes was relatively straightforward since these areas did not intersect. Further validation is planned, with data acquisition over the three study sites slated for the summer of 2016.

In this paper, we report results of a study aimed at assessing the potential for using X-band single-pass radar interferometric coherence for forest canopy height estimation. We use datasets from the Tandem-X satellite pair collected over Canadian forest test sites, where supporting lidar data are available for validation. We first employ dual-copolarized modes to assess the potential of polarimetric interferometry for forest canopy height retrieval. We show that for this forest type, single polarization modes have better properties, including much improved spatial coverage. We develop a new algorithm for single polarization data and validate the canopy height products against lidar. We then extend the canopy height product over a mosaic of multiple swaths to demonstrate the potential for very wide area forest height mapping using radar coherence.

In this paper we report results of a study aimed at assessing the potential for using the standard mode of TanDEM-X (single polarization HH, single baseline and single-pass radar interferometry) to estimate forest canopy heights over two Canadian forest test sites, where supporting LIDAR data is available for validation. We first apply two coherence correction factors to TanDEM-X data, one for signal-to-noise ratio (SNR) based on noise statistics from TanDEM-X metadata and the other on a system correction based on bare surface reference scattering. We then use coherence amplitude of the interferogram and interferometric wavenumber k(z) to estimate forest canopy heights. The radar heights are validated against a LIDAR CHM (Canopy Height Model) reference. A simple linear scaling of the TanDEM-X height products based on a relatively small LIDAR calibration patch is used to extend the canopy height products over a mosaic of multiple swaths for very wide area forest height mapping using radar coherence.

The forest biome is vital to the health of the earth. Canada and the United States have a combined forest area of 4.68 Mkm(2). The monitoring of these forest resources has become increasingly complex. Hyperspectral remote sensing can provide a wealth of improved information products to land managers to make more informed decisions. Research in this area has demonstrated that hyperspectral remote sensing can be used to create more accurate products for forest inventory (major forest species), forest health, foliar biochemistry, biomass, and aboveground carbon. Operationally there is a requirement for a mix of airborne and satellite approaches. This paper surveys some methods and results in hyperspectral sensing of forests and discusses the implications for space initiatives with hyperspectral sensing

SUMMARYA 13.3 GHz scatterometer and a nadir-viewing radiometer measuring reflected radiance in the LANDSAT bands were flown simultaneously over a number of fields near Ottawa containing eleven different forage crops. The purpose of the experiment was to compare the relative ability of the optical and microwave sensors to differentiate the various crop types, and to investigate the advantages of combining optical and microwave measurements for crop discrimination.The coherent illumination and detection inherent in the scatterometer operation results in analysis techniques and statistical uncertainties which are quite different from those associated with the more traditional optical techniques. The data reduction procedure and resulting uncertainties are discussed in some detail.Confusion matrices were calculated for both the optical and the microwave data. Microwave and optical sensors provide complementary information which, when combined, permit the most accurate (88.4 ± 8.0 (s.d.) % for 9 classes) class...

Clustering (and classification) among other approaches of land-cover type discrimination for Polarimetric SAR (Pol-SAR) data often explicitly or implicitly assume a lot about the shape of the clusters (or the classes, in the case of classification). For example, this is an issue for Pol-SAR classification methods [1,2,3] that initialize clusters in decomposition parameter feature spaces [4], subsequently refining the clusters by Wishart moving-means iterations in coherency matrix (T3) space. Indeed, using the means as cluster (or class) representatives can be successful, provided that clusters in the data are compact, well separated, and convex. However, highly nonlinear features and unusually shaped clusters are often obtained when dealing with PolSAR data. To address this issue we present a data-driven hierarchical clustering technique. This we demonstrate for forest-type discrimination purposes with a multi-temporal Radarsat-2 sandwich.

This paper gives results of our assessment of the potential for using Radarsat Constellation Mission (RCM) C-band compact polarimetry (compact-pol) for detecting historical forest fire scars. We first summarize the compact-pol decomposition theory we developed for retrieving useful geophysical parameters from compact-pol data. We then demonstrate a combination of time series filtering and spatial filtering to reduce speckle noise in the geophysical parameters. Next we describe a rule-based classifier and show an application example based on a time series of simulated compact-pol data from Radarsat-2 Fine Quad-pol (FQ) mode to detect a 10-year old fire scar in our study site. Our study results showed that even though there was a loss of polarimetric information through projection of a complex scattering matrix of quad-pol data on a single-pixel level, the compact-pol mode was capable of maintaining important polarimetric information and detecting the test forest fire scar. Finally we look at the effect of non-circular transmit polarization on key decomposition parameters and discuss the effects of imperfect transmit polarization on classification performance.

In this paper we analyze the effects of polarization, environmental conditions and forest structure upon the backscatter response of forested stands. This analysis is based upon a time series of ALOS-1 PALSAR images acquired over our study site in Xunke County, Heilongjiang Province, China. Based on six scenes, we analyzed the polarization and environment conditions on the forest stands. Backscatter coefficients of HV channel had a greater dynamic range than HH channel. HV channel was less influenced by weather and wind speed conditions. Our observations found canopy density greatly influenced the forest stand backscatter. Backscatter coefficient showed weak correlations to canopy density, mean tree height and mean diameter at breast height (DBH). Correlations were much stronger when the forest stands were grouped with canopy density or mean tree height. Before grouping the highest correlation coefficient between backscatter coefficients and tree height was 0.377, the value for HV image acquired on August 07, 2007. After grouping these forest stands by canopy density, the correlation was 0.95 to tree height. We also analyzed the correlations with two different tree species, and obtained similar results.

Compact polarimetry (compact-pol) architecture is a new hybrid mode for synthetic aperture radar (SAR) and is proposed for the future Canadian Radarsat Constellation Mission (RCM) and the future Japanese ALOS-2 mission. The Compact-pol mode transmits single circular polarization (left/right) and receives simultaneous coherent orthogonal linear polarizations. In this paper, we use Radarsat-2 C-band quad polarization (quad-pol) data to simulate RCM compactpol data, introduce useful compact-pol parameters using compact-pol decomposition theories, and show that the compact-pol mode is capable of maintaining important polarimetric information on forest structure and changes, even though there is the loss of information through projection of the complex scattering matrix of quad-pol on a single-pixel level. Examples are given to illustrate how to use the compact decomposition parameters for forestry applications.

The papers in this special issue represent the state of the art on forest modeling and mapping techniques to provide up-to-date, consistent information on forests at different scales, ranging from local, regional through to transnational and global scales. The papers can be broadly categorized in two categories: those dealing with passive, optical remote sensing applications with a particular focu...

In this paper we report results of a study aimed at assessing the potential for using X-band polarimetric interferometry for forest height estimation in high latitude boreal and temperate rainforest environments. We first summarize the data sets available from the Tandem-X satellite pair for our two main test sites before deriving a new height estimation algorithm and finally comparing radar derived heights against reference lidar canopy height data for the two scenes.

With increasing CO2 in the atmosphere due to fossil fuel burning, there is a need to quantitatively measure the aboveground carbon in forests. The best remote sensing sensors for this task in Canada are hyperspectral sensors to obtain major forest species, and lidar to measure tree height (H). The Greater Victoria Watershed District on Vancouver Island was selected as a test site and imaged with airborne AVIRIS 4m data and AISA 2m data. Fifty-four ground plots provided excellent ground reference data. Knowing the species, tree heights, and allometric equations relating to these species permits us to determine the aboveground carbon. This paper discusses these measurements, and the variation in carbon estimates due to errors of tree height and species classification.

In this paper we apply the ideas of target decomposition theory to compact polarimetry and use them to investigate forest applications at L-Band. We consider eigenvector methods (H/alpha), model based decompositions (RVOG) and coherent change detection (CCD) and show that compact modes are capable of maintaining important polarimetric information on forest structure and changes.