Multi-temporal SAR Data Research Articles

Delineating flood risk areas in Greater Dhaka of Bangladesh using geoinformatics

This paper illustrates the development of flood hazard and risk maps in Greater Dhaka of Bangladesh using geoinformatics. Multi-temporal RADARSAT SAR and GIS data were employed to delineate flood hazard and risk areas for the 1998 historical flood. Flood-affected frequency and flood depth were estimated from multi-date SAR data and considered as hydrologic parameters for the evaluation of flood hazard. Using land-cover, gemorphic units and elevation data as thematic components, flood hazard maps were created by considering the interactive effect of flood frequency and flood water depth concurrently. Analysis revealed that a major portion of Greater Dhaka was exposed to high to very high hazard zones while a smaller portion (2.72%) was free from the potential flood hazard. Flood risk map according to administrative division showed that 75.35% of Greater Dhaka was within medium to very high risk areas of which 53.39% of areas are believed to be fully urbanized by the year 2010.

Mapping of flood dynamics and spatial distribution of vegetation in the Amazon floodplain using multitemporal SAR data

This paper presents the use of time series of SAR images to map the flood temporal dynamics and the spatial distribution of vegetation over a large Amazonian floodplain. The region under study (3500 km 2) presents a diversity of landscape units with open lakes, bogs, large meadows, savannahs, alluvial forests and terra firma forest, covered by 21 images acquired by J-ERS between 1993 and 1997. Ground data include in situ observations of vegetation structure and flood extent as well as water level records. Image analysis demonstrates that temporal variations of the radar backscatter can be used to monitor efficiently the flood extent regardless of the landscape units. Also, analysis of the backscatter temporal variation greatly reduces the confusion between smooth surfaces (e.g. open water bodies, bare soils) inherent to L-band backscatter. The mapping method is based on decision rules over two decision variables: 1) the mean backscatter coefficient computed over the whole time series; 2) the total change computed using an “Absolute Change” estimator. The first variable provides classification into rough vegetation types while the second variable yields a direct estimate of the intensity of change that is related to flood dynamics. The classifier is first applied to the whole time series to map the maximum and minimum flood extent by defining 3 flood conditions: never flooded (NF); occasionally flooded (OF); permanently flooded (PF). It also furnishes the broad land cover type: open water/bare soils/low vegetation/forest. The accuracy of the flood extent mapping shows a kappa value of 0.82. Then, the classifier is run iteratively on the OF pixels to monitor flood stages during which the occasionally flooded areas get submerged. The mapping accuracy is assessed on one intermediate flood stage, showing a precision in excess of 90%. The importance of the time sampling for flood mapping is discussed along with the influence of SAR backscatter accuracy and the number of images. Then general guidelines for floodplain mapping are presented. By combining water level reports with maps of different flood stages the flooding pattern can be retrieved along with the vegetation succession processes. It is shown that the spatial distribution of vegetation communities is governed by flood stress and can be modelled as a function of the mean annual exposure to floods.

Dry snow discrimination in alpine areas from multi-frequency and multi-temporal SAR data

A new method is presented to discriminate dry snow from bare earth in alpine regions. Because of the wide variety of alpine environments, the scene under study is first segmented into surface and forested area classes from L-band summer polarimetric SAR data. Dry snow is then discriminated over each class, using adaptive methods. A new polarimetric multi-temporal optimisation procedure, named polarimetic contrast variation enhancement (PCVE), is proposed to increase the slight polarimetric contrast because of the presence of snow over surfaces. Snow-covered forests are discriminated using performing polarimetric indicators resulting from a decomposition of incoherent matrix representations. The effectiveness of this method is demonstrated over two different alpine test sites, using SIR-C L- and C-bands polarimetric SAR data.

Multi-temporal JERS SAR data in boreal forest biomass mapping

Multi-temporal JERS SAR data were studied for forest biomass mapping. The study site was located in South-eastern Finland in Ruokolahti. Pre-processing of JERS SAR data included ortho-rectification and radiometric normalization of topographic effects. In single-date regression analysis between backscatter amplitude and stem volume, summer scenes from July to October produced correlation coefficients ( r) between 0.63 and 0.81. Backscatter level and the slope of the (linear) regression line were stable from scene to scene. Winter scenes acquired in very cold and dry winter conditions had a very low correlation. One winter scene acquired in conditions where snow is not completely frozen produced a correlation coefficient similar to summer scenes. Multivariate regression analysis with a 6-date JERS SAR dataset produced correlation coefficient of 0.85. A combined JERS–optical regression analysis improved the correlation coefficient to 0.89 and also alleviated the saturation, which affects both SAR and optical data. The stability of the regression results in summer scenes suggests that a simple constant model could be used in wide-area forest biomass mapping if accuracy requirements are low and if biomass estimates are aggregated to large areal units.

Multi-scale analysis of intrinsic soil factors from SAR-based mapping of drying rates

Intrinsic soil factors affect and are affected by the spatial variation of soil properties. Therefore, intrinsic soil factors may both characterize and serve as an indicator for soil taxonomy. Difficulties in inferring intrinsic soil properties hamper attempts to assess their variability, on both local and regional/broad scales. Radar remote sensing might facilitate a breakthrough in this field, due to its sensitivity to the soil water content. In this research, a raster Geographic Information System (GIS) methodology is developed for combining multi-temporal ERS-2 SAR and Landsat TM data, which allows the estimation of drying rate patterns in bare soil surfaces. The drying rates provide further indication about intrinsic soil properties. The multi-scale behaviour of soil-drying rates is described using the richness–area curves and characteristic curves are determined to four soil formations typical to a climatic gradient between Mediterranean and semi-arid environments in Israel. To the best of our knowledge, this is one of the first attempts to document the effect of intrinsic soil factors on the soil system at the regional scale. The results achieved here demonstrate the connection between drying rates, richness–area variation and soil hydraulic conductivity of the four soil formations.

Assessment of post-flooding conditions of rice fields with multi-temporal satellite SAR data

The temporal characteristics of radar response from rice fields were analysed throughout the 1999 summer flooding season in the agricultural area of the central Korean Peninsula. Multiple scenes of Radarsat SAR images were acquired before, during, and after the flood inundation. From the interpretation of colour composite imagery of the multi-temporal SAR data, as well as from the temporal profiles of radar backscatter, three types of rice fields were separable according to flooding and post-flooding recovery conditions. The actual flooding damage to the standing rice crop depends on the duration of the flood and on subsequent recovery status. We found that image data acquired during and after flooding is necessary to assess accurately flood damage to rice fields.

Towards an operational system for regional-scale rice yield estimation using a time-series of Radarsat ScanSAR images

This paper demonstrates that Radarsat ScanSAR data can be an important data source of radar remote sensing for monitoring crop systems and estimation of rice yield in large areas in tropical and sub-tropical regions. Experiments were carried out to show the effectiveness of Radarsat ScanSAR data for rice yield estimation in the whole province of Guangdong, South China. A methodology was developed to deal with a series of issues in extracting rice information from the ScanSAR data, such as topographic influences, levels of agro-management, irregular distribution of paddy fields and different rice cropping systems. A model was provided for rice yield estimation based on the relationship between the backscatter coefficient of multi-temporal SAR data and the biomass of rice. The study indicates that the whole procedure can become a low-cost and convenient operational system for large-scale rice yield estimation which is difficult for conventional methods.

Investigating the potential for soil moisture and surface roughness monitoring in drylands using ERS SAR data

Multitemporal ERS-1 and ERS-2 SAR data were acquired for northern Jordan between 1995 and 1997 to investigate changes in the backscatter coefficients of a range of typical desert land surfaces. The changes in backscatter found were ascribed to variations in surface soil moisture, and changes in surface roughness caused by a range of natural and anthropogenic factors. Data collected from monitored sites were input into the Integral Equation Model (IEM). The model outputs were strongly correlated with observed backscatter coefficients (r 2=0.84). The results show that the successful monitoring of soil moisture in these environments is strongly dependent on the surface roughness. On surfaces with RMS height 0.5 cm, the sensitivity of the backscatter coefficient to changes in surface microtopography did not allow accurate soil moisture estimation. Microtopographic change on rougher surfaces has less influence on the backscatter coefficient, and the probability of soil moisture estimation from SAR imagery is g...

Investigating the potential for soil moisture and surface roughness monitoring in drylands using ERS SAR data

Multitemporal ERS-1 and ERS-2 SAR data were acquired for northern Jordan between 1995 and 1997 to investigate changes in the backscatter coefficients of a range of typical desert land surfaces. The changes in backscatter found were ascribed to variations in surface soil moisture, and changes in surface roughness caused by a range of natural and anthropogenic factors. Data collected from monitored sites were input into the Integral Equation Model (IEM). The model outputs were strongly correlated with observed backscatter coefficients (r 2=0.84). The results show that the successful monitoring of soil moisture in these environments is strongly dependent on the surface roughness. On surfaces with RMS height 0.5 cm, the sensitivity of the backscatter coefficient to changes in surface microtopography did not allow accurate soil moisture estimation. Microtopographic change on rougher surfaces has less influence on the backscatter coefficient, and the probability of soil moisture estimation from SAR imagery is greater. These results indicate that knowledge of the surface conditions (both in terms of surface roughness and geomorphology) is essential for accurate soil moisture monitoring, whether in a research or operational context. The potential benefits of these findings are discussed in the context of the Jordan Badia Research and Development Project.

Monitoring of new plantation development in tropical rain forests using JERS-1 SAR data

Abstract The authors studied the applicability of multi temporal SAR data obtained from JERS-1 SAR for monitoring the change in tropical rain forest conditions due to plantation development. The test site was the southern part of Sumatra Island, Indonesia. A total of seven JERS-1 SAR and one ERS-2 SAR data from 1992 to 1997 were analyzed together with two optical data, MOS-1 MESSR in 1990 and SPOT HRV in 1997. The result of this study verified that the backscatter change in multi temporal JERS-1 SAR data can be used effectively to monitor the process of plantation development in the test site, which accompanies the burning of felled tree trunks after deforestation, one of the causes for forest fire and serious smoke.

Comparison of Algorithms for Classifying Swedish Landcover Using Landsat TM and ERS-1 SAR Data

Sixteen landcover classes in a representative Swedish environment were analyzed and classified using one Landsat TM scene and seven ERS-1 SAR.PRI images acquired during 1993. Spectral and backscattering signature separabilities are analyzed using the Jeffries–Matusita distance measure to determine which combinations of channels/images contained the most information. Maximum likelihood, sequential maximum a posteriori (SMAP, a Bayesian image segmentation algorithm), and back propagation neural network classification algorithms were applied and their performances evaluated. Results of the separability analyses indicated that the multitemporal SAR data contained more separable landcover information than did the multispectral TM data; the highest separabilities were achieved when the TM and SAR data were combined. Classification accuracy evaluation results indicate that the SMAP algorithm outperformed the maximum likelihood algorithm which, in turn, outperformed the neural network algorithm. The best KAPPA values, using combined data, were 0.495 for SMAP, 0.445 for maximum likelihood, and 0.432 for neural network. Corresponding overall accuracy values were 57.1%, 52.4%, and 51.2%, respectively. A comparison between lumped crop area statistics with areal sums calculated from the classified satellite data gave the highest correspondence where the SMAP algorithm was used, followed by the maximum likelihood and neural network algorithms. Based on our application, we can therefore confirm the value of a multisource optical/SAR approach for analyzing landcover and the improvements to classification achieved using the SMAP algorithm.

Multitemporal ERS-1 SAR Data for Crop Classification: A Sequential-Masking Approach

RÉSUMÉLa technique de masquage séquentiel basée sur des procédures de photo-interprétation permet de différencier les caractéristiques des images en utilisant une série d'images multidates. Dans cette étude, un jeu de neuf images radar ERS-1 est analysé au moyen de cette technique pour déterminer les dates les plus hâtives à partir desquelles il est possible d'identifier les différents types de culture dans une région agricole du sud de l'Ontario, Canada. Les profils temporels de rétrodiffusion radar des cultures ont été générés à partir d'images radar étalonnées. À partir de ces profils temporels de rétrodiffusion, on a réalisé des classifications par champ individuel en appliquant la technique de masquage séquentiel aux données radar multidates acquises en début de saison et en mi-saison. Les résultats démontrent qu'à l'aide de seulement trois images acquises le 31 mai, le 16 juin et le 5 juillet, il est possible de différencier le blé d'hiver, luzerne/foin, orge/avoine, le soja et le maïs avec une précision globale de validation de 88,5% et un coefficient Kappa de 0,85.

Multi-temporal analysis of ERS-1 and EMISAR C-band VV backscattering properties of forest and lake surfaces in the NOPEX region

Multi-temporal ERS-1 SAR data, consisting of seven acquisitions in 1994 and 1995, were analyzed to determine the potential of C-band VV polarization backscattering data to discriminate different surface types within the patchy boreal landscape of the NOPEX area. Based on an aerial photo analysis, four classes of forest density were distinguished. For independent comparison, also one fully polarimetric EMISAR C-band image, including the VV polarization signatures was analyzed. It was found that the differences between the classes were quite consistent throughout the seasons and therefore probably significant, at least in a relative sense. The same images were also studied to determine the sensitivity of backscattering properties of lake surfaces to regional winds. For this purpose the backscattering values of five lakes within the study area were analyzed. It was found that the five lake surfaces behaved very similar, except for one observation date. This behavior, therefore, seems to be dominated by regional scale wind fields, although local scale wind fields also may have an effect. Since regional and local scale wind fields are difficult to observe in a synoptic way, the use of radar might have potential for monitoring such wind fields in the Swedish boreal region, which is characterized by the occurrence of many lakes.

Analysis of Multi-Temporal ERS-1 SAR Data of Subarctic Tundra and Forest in the Northern Hudson Bay Lowland and Implications for Climate Studies

RÉSUMÉDes données satellite radar ERS-1 out été utilisées pour analyser les variations annuelles de la rétrodiffusion dans la tundra sub-arctique et la forêt dans la partie nord des basses terres de la Baie d'Hudson (HBL) près de Churchill, Manitoba. On utilise des données de terrain relatives aux conditions physiques, écologiques et météorologiques acquises dans le cadre de cette étude afin d'interpréter les variations de la rétrodiffusion à partir de cinq sites représentant des types de terrain caractéristiques des terres basses de la Baie d'Hudson (fen, plage soulevée, arbustes de saule/bouleau, forêt ouverte et lac). Les plus grandes variations de rétrodiffusion observées à l'intérieur d'un même type de terrain au cours d'un cycle annuel entier (août 1992 à août 1993) sont associées à des périodes de gel et dégel et à l'humidité de surface (humidité du sol et quantité d'eau dormante). Les différences des valeurs de rétrodiffusion entre les sites dépendent de la nature de la surface qui détermine l'ampleur du processus de rétrodiffusion (volume ou surface). L'impact de l'utilisation des données satellite multidate sur les recherches en cours dans le domaine des études climatiques dans la région des basses terres est mis en relief.

Temporal and spatial characteristics of irrigated rice in JERS-1 L-band SAR data

Multi-temporal JERS-1 SAR (Japanese Earth Resource Satellite 1 Synthetic Aperture Radar) data have been used in this study to investigate the temporal and spatial characteristics of irrigated rice in LHH-band SAR data. The two test sites chosen represent areas that are characterized by manual and mechanical planting practices that turn out to have a significant effect on the L-band backscatter. The results show a clear relationship between L-band backscatter and plant growth in manually planted areas. A dynamic range of about 8dB has been measured with no apparent saturation of the signal. Areas in which mechanical planting is practiced, however, display a more complex relationship where the radar response is strongly dependent on parameters related to the spatial distribution of the plants. The backscatter displays a strong dependency on planting orientation and plant spacing, where fields planted in the radar illumination direction with a certain plant spacing display extreme dynamic ranges of up to more than 20dB due to resonant scattering. Fields planted in other directions generally experience attenuation of the signal and dynamic ranges do not change by more than a few decibels. Despite these artifacts, however, a correlation between plant growth and backscatter has been observed for all fields, independent of plant spacing and field orientation.

Soil Moisture Measurement using Active Microwave Remote Sensing

Ground-base verification experiments for the soil moisture detection by spaceborne SARs (E ERS-1 and J ERS-1) were carried out. Backscattering coefficients obtained from satellite SARs were compared with those measured by ground-base scatterometer system and ground truth data of soil moisture, and the relationship between backscattering coefficients measured by SARs and soil moisture content was examined. Major results obtained from the study are as follows; (i) Backscattering coefficients measured by E ERS-1 AMI are not more sensitive to surface conditions than those by the scatterometer and have a high correlation with soil moisture content for the soil moisture range of less than 40 correlation with the saturation degree of soil moisture, but low sensitivity to it. (iii) Each land cover has its own seasonal change, which shows the possibility to classify land cover types from multitemporal SAR data. (iv) Using SAR data with visible satellite data, soil moisture distribution maps are made up and validated with qualitative observations.

能動型マイクロ波リモートセンシングによる表層土壌水分計測 Ｉ． 室内実験

全球スケールでの合成開口レーダ(SAR)によるモニタリングが実用化段階に入り,水文学の領域では表層土壌水分量と積雪内水分量の定量的把握が期待されている.本研究では能動型マイクロ波リモートセンシングによる表層土壌水分量に焦点を当て,地上計測用マイクロ波Cバンド散乱計を製作し,これを用いて重要な各種特性について室内実験を行なった.まずアンテナ特性などのシステム特性を電波暗室で調べる実験を行なった.さらに,この実験システムを用いて,マイクロ波土壌侵入深,土壌水分感度,入射角依存性,地表面粗度の影響について室内実験を行ない,以下の点が明らかとなった.(i)Cバンドマイクロ波の土壌への侵入深は,それが最も大きくなる乾燥時の鉛直入射状態でも10cm程度であり,SARの様なサイドルッキングレーダーでは土壌侵入深はより浅くなる.(ii)散乱計で観測される後方散乱係数と表層土壌水分量の関係を定式化する場合に実用的な表層土壌深は2.5cmである.(iii)後方散乱係数の入射角依存性は入射角が小さいほど顕著である.ただし,本室内実験システムでは30°以上の入射角については良い実験結果が得られなかった.(iv)HV偏波よりもHH偏波の方が後方散乱係数の土壌水分感度が高い.(v)対象土壌面が滑面と粗面とでは,後方散乱係数に非常に大きな差が生じる.

Seasonal dynamics of C-band backscatter of boreal forests with applications to biomass and soil moisture estimation

The seasonal changes of the C-band backscattering properties of boreal forests are investigated by applying 1) a semiempirical forest backscattering model and 2) multitemporal ERS-1 SAR data from two test areas in Finland. The semiempirical modeling of forest canopy volume backscattering and extinction properties is based on high-resolution data from the authors' ranging scatterometer HUTSCAT. The response of ERS-1 SAR to forest stem volume (biomass) and other forest characteristics is investigated by employing the National Forest Inventory sample plots, stand-wise forest inventory data and LANDSAT- and SPOT-based digital land use maps. The results show that the correlation between the backscattering coefficient and forest stem volume (biomass) varies from positive to negative depending on canopy and soil moisture. Additionally, the seasonal snow cover and soil freezing/thawing effects cause drastic changes in the radar response. A novel method for the estimation of forest stem volume (biomass) is introduced. This technique is based on the use of: 1) multitemporal ERS-1 SAR data; 2) reference sample plot information; and 3) the semiempirical backscattering model. It is shown that the multitemporal ERS-1 SAR images can be successfully used for estimating the forest stem volume. The effects of soil moisture variations to ERS-1 SAR results have been analyzed using hydrological soil moisture model and in situ data. The results indicate that the semiempirical model can he used for predicting the soil and canopy moisture variations in ERS-1 images.

Differentiating methane source areas in Arctic environments with multitemporal ERS-1 SAR data

An assessment using ERS-1 SAR data to differentiate methane source (wetland) and nonsource (nonwetland) areas was undertaken based on radar backscatter modeling and empirical observations of 24 scenes collected over Barrow, AK, in 1991 and 1992. Differences in backscatter between source and nonsource areas were dependent on surface hydrology and air temperature. Differential freezing of surface materials on daily to seasonal time scales greatly enhanced the separability of wetlands and nonwetlands with ERS-1 SAR. Radar return for nonwetlands decreased dramatically whereas backscatter from wetlands decreased little when freezing air temperatures coincided with the SAR overpass. Maximum separability between wetlands and nonwetlands, as determined from observed and modeled radar backscatter, were the result of changes in the dielectric constant of the plant and surface materials with phase change during freezing. This study has indicated the need to consider air temperature at the time of acquisition in selecting ERS-1 SAR scenes for differentiating methane source and nonsource areas.

Monitoring of rice crop growth from space using the ERS-1 C-band SAR

This communication presents preliminary results of the first attempt to monitor rice crop growth from space from transplanting to harvesting with an active microwave instrument using ERS-1 C-band SAR. The variation of /spl sigma//sup 0/ for a test rice field in the Akita Prefectural College of Agriculture obtained from multitemporal SAR data is evaluated by comparing with ground truth data of the rice crop condition. The feasibility is shown of estimating the number of days after transplanting and hence the rice crop growth stage from the /spl sigma//sup 0/ measurement. Correlation analysis suggests a close relationship between /spl sigma//sup 0/ and the two dimensional extent of the rice crops. >