S. Saatchi

Large seasonal swings in leaf area of Amazon rainforests

Despite early speculation to the contrary, all tropical forests studied to date display seasonal variations in the presence of new leaves, flowers, and fruits. Past studies were focused on the timing of phenological events and their cues but not on the accompanying changes in leaf area that regulate vegetation–atmosphere exchanges of energy, momentum, and mass. Here we report, from analysis of 5 years of recent satellite data, seasonal swings in green leaf area of ≈25% in a majority of the Amazon rainforests. This seasonal cycle is timed to the seasonality of solar radiation in a manner that is suggestive of anticipatory and opportunistic patterns of net leaf flushing during the early to mid part of the light-rich dry season and net leaf abscission during the cloudy wet season. These seasonal swings in leaf area may be critical to initiation of the transition from dry to wet season, seasonal carbon balance between photosynthetic gains and respiratory losses, and litterfall nutrient cycling in moist tropical forests.

The use of decision tree and multiscale texture for classification of JERS-1 SAR data over tropical forest

The objective of this paper is to study the use of a decision tree classifier and multiscale texture measures to extract thematic information on the tropical vegetation cover from the Global Rain Forest Mapping (GRFM) JERS-1 SAR mosaics. We focus our study on a coastal region of Gabon, which has a variety of land cover types common to most tropical regions. A decision tree classifier does not assume a particular probability density distribution of the input data, and is thus well adapted for SAR image classification. A total of seven features, including wavelet-based multiscale texture measures (at scales of 200, 400, and 800 m) and multiscale multitemporal amplitude data (two dates at scales 100 and 400 m), are used to discriminate the land cover classes of interest. Among these layers, the best features for separating classes are found by constructing exploratory decision trees from various feature combinations. The decision tree structure stability is then investigated by interchanging the role of the training samples for decision tree growth and testing. We show that the construction of exploratory decision trees can improve the classification results. The analysis also proves that the radar backscatter amplitude is important for separating basic land cover categories such as savannas, forests, and flooded vegetation. Texture is found to be useful for refining flooded vegetation classes. Temporal information from SAR images of two different dates is explicitly used in the decision tree structure to identify swamps and temporarily flooded vegetation.

The Global Rain Forest Mapping project— a review

The Global Rain Forest Mapping (GRFM) project is an international endeavour led by the National Space Development Agency of Japan (NASDA), with the aim of producing spatially and temporally contiguous Synthetic Aperture Radar (SAR) data sets over the tropical belt on the Earth by use of the JERS-1 L-band SAR, through the generation of semi-continental, 100 m resolution, image mosaics. The GRFM project relies on extensive collaboration with the National Aeronautics and Space Administration (NASA), the Joint Research Centre of the European Commission (JRC) and the Japanese Ministry of International Trade and Industry (MITI) for data acquisition, processing, validation and product generation. A science programme is underway in parallel with product generation. This involves the agencies mentioned above, as well as a large number of international organizations, universities and individuals to perform field activities and data analysis at different levels. The GRFM project was initiated in 1995 and, through a dedicated data acquisition policy by NASDA, data acquisitions could be completed within a 1.5-year period, resulting in a spatially and temporally homogeneous coverage to encompass the entire Amazon Basin from the Atlantic to the Pacific; Central America up to the Yucatan Peninsular in Mexico; equatorial Africa from Madagascar and Kenya in the east to Sierra Leone in the west; and south-east Asia, including Papua New Guinea and northern Australia. Over the Amazon and Congo river basins, the project aimed to provide complete cover at two different seasons, featuring the basins at high and low water. In total, the GRFM acquisitions comprise some 13000 SAR scenes, which are currently in the course of being processed and compiled into image mosaics. In March 1999, SAR mosaics over the Amazon Basin (one out of two seasonal coverages) and equatorial Africa (both seasonal coverages) were completed; the data are available on CD-ROM and, at a coarser resolution, via the Internet. Coverage of the second-season Amazon and Central America will be completed during 1999, with the south-east Asian data sets following thereafter. All data are being provided free of charge to the international science community for research and educational purposes.

Mapping deforestation and land use in amazon rainforest by using SIR-C imagery☆

Abstract In this paper, the potential use of spaceborne polarimetric synthetic aperture radar (SAR) data in mapping landcover types and monitoring deforestation in tropics is studied. Here, the emphasis is placed on several clearing practices and forest regeneration that can be characterized by using the sensitivity of SAR channels to vegetation biomass and canopy structure. A supervised Bayesian classifier designed for SAR signal statistics is employed to separate five classes: primary forest, secondary forest, pasture-crops, quebradao, and disturbed forest. The L- and C-band polarimetric SAR data acquired during the shuttle imaging radar-C (SIR-C)/X-SAR space-shuttle mission in 1994 are used as input data to the classifier. The results are verified by field observation and comparison with the Landsat data acquired in August of 1994. The SAR data can delineate these five classes with approximately 72% accuracy. The confusion arises when separating old secondary forests from primary forest and the young ones from pasture-crops. It is shown that Landsat and SAR data carry complementary information about the vegetation structure that, when used in synergism, may increase the classification accuracy over secondary forest regrowth. When the number of land-cover types was reduced to three classes including primary forest, pasture-crops, and regrowth-disturbed forest, the accuracy of classification increased to 87%. A dimensionality analysis of the classifier showed that the accuracy can be further improved to 92% by reducing the feature space to L-band HH and HV channels. Comparison of SIR-C data acquired in April (wet period) and October (dry period) indicates that multi-temporal data can be used for monitoring deforestation; however, the data acquired during the wet season are not suitable for accurate land-cover classification.

On the detection of Faraday rotation in linearly polarized L-band SAR backscatter signatures

The potentially measurable effects of Faraday rotation on linearly polarized backscatter measurements from space are addressed. Single-polarized, dual-polarized, and quad-polarized backscatter measurements subject to Faraday rotation are first modeled. Then, the impacts are assessed using L-band polarimetric synthetic aperture radar (SAR) data. Due to Faraday rotation, the received signal will include other polarization characteristics of the surface, which may be detectable under certain conditions. Model results are used to suggest data characteristics that will reveal the presence of Faraday rotation in a given single-polarized, dual-polarized, or quad-polarized L-band SAR dataset, provided the user can identify scatterers within the scene whose general behavior is known or can compare the data to another, similar dataset with zero Faraday rotation. The data characteristics found to be most sensitive to a small amount of Faraday rotation (i.e., a one-way rotation <20/spl deg/) are the cross-pol backscatter [/spl sigma//spl deg/(HV)] and the like-to-cross-pol correlation [e.g., /spl rho/(HHHV/sup */)]. For a diverse, but representative, set of natural terrain, the level of distortion across a range of backscatter measures is shown to be acceptable (i.e., minimal) for one-way Faraday rotations of less than 5/spl deg/, and 3/spl deg/ if the radiometric uncertainty in the HV backscatter is specified to be less than 0.5 dB.

Estimation of crown and stem water content and biomass of boreal forest using polarimetric SAR imagery

Characterization of boreal forests in ecosystem models requires temporal and spatial distributions of water content and biomass over local and regional scales. The authors report on the use of a semi-empirical algorithm for deriving these parameters from polarimetric synthetic aperture radar (SAR) measurements. The algorithm is based on a two layer radar backscatter model that stratifies the forest canopy into crown and stem layers and separates the structural and biometric attributes of forest stands. The structural parameters are estimated by training the model with SAR image data over dominant coniferous and deciduous stands in the boreal forest such as jack pine, black spruce, and aspen. The algorithm is then applied on AIRSAR images collected during the Boreal Ecosystem Atmospheric Study (BOREAS) over the boreal forest of Canada. The results are verified using biometry measurements during BOREAS-intensive field campaigns. Field data relating the water content of tree components to dry biomass are used to modify the coefficients of the algorithm for crown and stem biomass. The algorithm was then applied over the entire image generating biomass maps. A set of 18 test sites within the imaged area was used to assess the accuracy of the biomass maps. The accuracy of biomass estimation is also investigated by choosing different combinations of polarization and frequency channels of the AIRSAR system. It is shown that polarimetric data from P-band and L-band channels provide similar accuracy for estimating the above-ground biomass for boreal forest types. In general, the use of P-band channels can provide better estimates of stem biomass, while L-band channels can estimate the crown biomass more accurately.

Global ice and land climate studies using scatterometer image data

Scatterometers have provided continuous synoptic microwave radar coverage of the Earth from space for nearly a decade. NASA launched three scatterometers: the current SeaWinds scatterometer onboard QuikSCAT (QSCAT, 13.4 GHz) launched in 1999; the NASA scatterometer (NSCAT, 14.0 GHz), which flew on the Japanese Space Agencys ADEOS-1 platform during 1996–1997; and the Seasat-A scatterometer system (SASS, 14.6 GHz), which flew in 1978. The European Space Agencys (ESA) 5.3-GHz scatterometer (ESCAT) has been carried onboard both the ERS-1 and ERS-2 satellites since 1991. properties, including the phase state, of a particular surface type. Varying response from the surface also results from different polarizations, viewing angles and orientations, and radar frequencies. The wide swath of scatterometers provides near daily global coverage at intrinsic sensor resolutions that are generally between 25–50 km.

Estimating Subcanopy Soil Moisture with RADAR

The subcanopy soil moisture of a boreal old jack pine forest stand is estimated using polarimetric L and P band airborne synthetic aperture radar (AIRSAR) data. Model simulations have shown that for this stand the principal scattering mechanism responsible for radar backscatter is the double-bounce mechanism between the tree trunks and the ground. The data to be used here were acquired during five flights from June to September 1994 as part of the Boreal Ecosystem-Atmosphere Study (BOREAS) project. The dielectric constants, or equivalently moisture contents, of the trunks and soil can change significantly during this period. To estimate these dynamic unknowns, parametric models of radar backscatter for the double-bounce mechanism are developed using a series of simulations of a numerical forest scattering model. A nonlinear optimization procedure is used to estimate the dielectric constants. Ground measurements of soil and trunk moisture content are used to validate the results. The trunk moisture content measurements are used to gain confidence that the respective estimation results are accurate enough not to corrupt the soil moisture estimation, which is the main focus of this paper. After conversion of the trunk moisture measurements to dielectric constants it is found that the estimated values are within 14% of the measurements. Owing to possible calibration uncertainties in the soil moisture measurements on the ground as well as in AIRSAR data, the variations rather than the absolute levels of the estimated soil moisture are considered. The results indicate that the estimated variations closely track the measurements. The worst case average estimated change differs by <1% volumetric soil moisture from that measured on the ground.

The Global Rain Forest Mapping Project JERS-1 radar mosaic of tropical Africa: development and product characterization aspects

The Global Rain Forest Mapping Project (GRFM) is an international collaborative effort initiated and managed by the National Space Development Agency of Japan (NASDA). The main goal of the project is to produce a high resolution wall-to-wall map of the entire tropical rain forest domain in four continents using the L-band SAR onboard the JERS-1 spacecraft. The processing phase, which entails the generation of wide area radar mosaics from the raw SAR data, was split according to the geographic area. In this paper, the focus is on the part related to Africa. The GRFM projects goal calls for the coverage of a continental scale area of several million km 2 using a sensor with the resolution of tens of meters. In the case of the African continent, this entails the assemblage of some 3900 high resolution SAR scenes into a bitemporal mosaic at 100 m pixel spacing and with known geometric accuracy. While this fact opens up an entire new perspective for vegetation mapping in the tropics, it presents a number of technical challenges. In this paper, we report on the solutions adopted in the GRFM Africa mosaic development and discuss some quantitative and qualitative aspects related to the characterization and validation of the GRFM products. In particular, the mosaic geolocation and its validation are discussed in detail. Indeed, the internal geometric consistency (subpixel accuracy in the coregistration of the two dates), and the absolute geolocation (residual mean squared error of 240 m with respect to ground control points) are key features of the GRFM Africa mosaic. Other important aspects that are discussed are the multiresolution decomposition approach, which allows for tracking the evolution of natural phenomena with scale; the internal semi-automatic radiometric calibration, which minimizes artifacts in the mosaic; and the thematic information content for vegetation mapping, which is illustrated by a few examples elaborated by visual interpretation. Experience gained so far indicates that the GRFM products constitute an important source of information for global environmental studies.

Microwave backscattering and emission model for grass canopies

Microwave radar and radiometer measurements of grasslands indicate a substantial reduction in sensor sensitivity to soil moisture in the presence of a thatch layer. When this layer is wet it masks changes in the underlying soil, making the canopy appear warm in the case of passive sensors (radiometer) and decreasing backscatter in the active case (scatterometer). A model for a grass canopy with thatch is presented in order to explain this behavior and for comparison with observations. The canopy model consists of three layers: grass, thatch, and the underlying soil. The grass blades are modeled by elongated elliptical discs and the thatch is modeled as a collection of disk shaped water droplets (i.e., the dry matter is neglected). The ground is homogeneous and flat. The distorted Born approximation is used to compute the radar cross section of this three layer canopy and the emissivity is computed from the radar cross section using the Peake formulation for the passive problem. Results are computed at L-band (1.4 GHz) and C-band (4.75 GHz) using canopy parameters (i.e., plant geometry, soil moisture, plant moisture, etc.) representative of Konza Prairie grasslands. The results are compared to C-band scatterometer measurements and L-band radiometer measurements at these grasslands. >