Danielle Hoja

DEM generation from ERS SAR shorelines compared to airborne crosstrack InSAR DEMs in the German Bight

Coastal regions like the intertidal area of the German Bight are continuously changing their topography due to morphodynamic processes. This fact has serious impact on the German waterways, requiring a huge amount of dragging with costs of about 8/spl times/10/sup 7/ DM per year. To investigate these morphodynamics and to reduce the costs, an up to date digital elevation model (DEM) is needed. The DEM currently used is determined by conventional vessel-mounted echo sounder measurements, which is difficult and expensive to update, especially in very shallow areas that are navigable only during extreme tides. In order to monitor morphodynamic changes and to continuously update the changes in the bottom topography, shoreline information extracted from synthetic aperture radar (SAR) images of the European remote sensing satellites (ERS-1/2) at different tidal levels are interpolated into a DEM. Therefore, the shorelines are extracted with a new developed method using a combination of wavelet and active contours. The corresponding height information is interpolated from tide gauge data. In a final step, from this height information a DEM is generated using a weighted mean interpolation in space and time. For validation purpose, the extracted shorelines and the interpolated DEM are compared to an independent. Generated DEM based on airborne crosstrack interferometric SAR images as well as on ground truth measurements. In total some 100 ERS SAR scenes from 1992 to 2000 are used to identify morphodynamical active areas in the German Bight.

A parametric scheme for ocean wave spectra retrieval from complex SAR data using prior information

A parametric inversion scheme for the retrieval of two dimensional ocean wave spectra from look cross spectra (LCS) acquired by spaceborne synthetic aperture radar (SAR) is presented. The scheme takes information about the spectral shape of different wave systems from a prior wave spectrum, while estimates for wavelength, waveheight and wave propagation direction are extracted from SAR cross spectra. The Partition Rescaling and Shift Algorithm (PARSA) is based on a partioning of a prior wave spectrum, e.g. taken from ocean wave models. For each ocean wave system a stochastic model is set up, which defines the probability that the propagation direction, the wavelength or the energy of the different wave systems deviate from the prior knowledge. The prescribed probabilities thereby quantify the confidence into the prior wave spectrum. Based on the probability models for the prior wave spectrum and the measured cross spectrum an optimal ocean wave spectrum is estimated using a maximum a posteriori approach. To solve the corresponding minimization problem the prior model is approximated with a multivariate Gaussian model. The optimization problem is solved with a Gauss-Newton method. The scheme is tested using both simulated cross spectra and reprocessed wave mode data acquired by the ERS-2 SAR. The reprocessed data are similar to the products, which will be available from the ENVISAT satellite launched in March 2002.

SAR measurements of ocean wind and wave fields in hurricanes

Spaceborne synthetic aperture radar (SAR) is still the only instrument providing directional information on surface wind and ocean waves on a global and continuous basis. From RADARSAT-1, many SAR images have been collected over the past years, which cover several hurricanes and allow investigation of the wind and wave fields under these extreme situations. A time-frame of 27 days of ERS-2 SAR wave mode data was processed, which covers several tropical cyclones in the Atlantic Ocean, of which Hurricane Edouard has been investigated in detail together with additional data available from scatterometers, buoys and weather centers. The wind fields and wave parameters are extracted from SAR imagery and compared to results of the numerical model output provided by the European Centre for Medium-Range Weather Forecast (ECMWF) and co-located ERS-2 scatterometer measurements. For each wave system, spectral parameters such as wavelength, wave propagation direction and wave age are calculated and compared to the numerical model output provided by ECMWF.

Global ocean wave measurements from ENVISAT ASAR data using a parametric inversion scheme

Two-dimensional ocean wave spectra are measured from ENVISAT ASAR wave mode cross spectra on a global scale. The measurement is performed using a parametric retrieval scheme, which makes use of prior information taken from numerical wave models. The Partition Rescale and Shift algorithm (PARSA) is based on a partitioning technique, which splits an a prior wave spectrum into its wave system components. Integral parameters of these systems, such as mean direction, mean wavelength, waveheight, and directional spreading are then adjusted iteratively to improve the consistency with the SAR observation. The method takes into account the full nonlinear SAR imaging process and uses a maximum a posteriori approach, which is based on statistical model quantifying the errors of the SAR imaging model, the SAR measurement, and the prior wave spectra. The method is applied to a global data set of ENVISAT ASAR data acquired during the CAL/VAL phase. The benefit of cross spectra compared to conventional symmetric image spectra is demonstrated.

Global analysis of ocean wave systems from SAR wave mode data

One of the major goals of the European Remote Sensing Satellites ERS-1 and ERS-2 were applications in ocean wave research and wave forecasting. For the first time, two-dimensional spectral information on the sea state globally, continuously, and in quasi real time are provided in the so-called SAR wave mode. To use not only the spectral but also the image information of these data, a global set of single look complex (SLC) synthetic aperture radar (SAR) images (imagettes) was processed from ERS-2 wave mode raw data, using the BSAR processor developed at DLR. SAR imagettes are used to analyze recently developed algorithms for wind and wave measurements. As the new European Satellite ENVISAT will provide cross spectra on a global and continuous basis, interest in the described techniques is growing. Two-dimensional ocean wave spectra are derived from SAR imagettes by quasi-linear inversion of the SAR imaging mechanism. Individual wave systems are detected by a so-called partitioning algorithm. The resulting wave parameters are analyzed and compared to model data. To take into account the complete nonlinear SAR imaging mechanisms, wave model spectra are simulated forward into cross spectra and compared to observations. Using this approach cases of significant inconsistencies between SAR and model data can be detected. With our 3 week data base in 1996 (about 30,000 imagettes) statistics are calculated. Finally, we derive daily maps showing areas with significant deviations between simulated and observed cross spectra.

Global tracking of swell with complex ERS-2 wave mode data

Complex synthetic aperture radar (SAR) data acquired by the European remote sensing satellite ERS-2 are used to analyze the propagation of swell on a global basis. The study is based on complex SAR wave mode imagettes, which were processed from ERS-2 SAR raw data using the DLR BSAR processor. Complex imagettes allow to apply the so called cross spectra technique globally for the first time. In contrast to conventional SAR intensity images, cross spectra provide information on ocean wave propagation without 180/spl deg/ ambiguity. A test data set of three weeks of complex imagettes (/spl ap/36000) is used to measure ocean swell. Distribution of swell is studied based on maps showing swell direction, wave height and wavelength. Temporal swell dynamics is analyzed using data acquisitions with small spatial distance and a time gap of /spl ap/12 h. Apart from providing information on ocean swell the study is a preparation for the coming ENVISAT era where cross spectra will be a standard product of the European Space Agency (ESA).

Detailed Damage Assessment After the Haiti Earthquake

In the post crisis phase a much more detailed analysis can be done with higher accuracy and less pressure of time compared to the general situation assessment of the rapid mapping process directly after the crisis. In this investigation the analysis is concentrated on the urban area of the capital town Port-au-Prince. In order to develop a service for detailed damage assessment, methods of (semi-)automatic change detection are used and compared, since up to now, damage information was mainly derived by visual interpretation. Any improvement in terms of accuracy and speed of analysis is of relevance to users in this context. The results of the different change detection algorithms achieved using the Haiti datasets are compared to each other and also to the database of the Haiti Action Plan for Reconstruction and Development (PDNA). The results are very promising although further improvements have to be made.

Use of the coherence of synthetic aperture radar cross spectra for ocean wave measurements

A noise model for synthetic aperture radar (SAR) look cross spectra (LCS) acquired over the ocean is proposed. The study is meant to contribute to the improvement of algorithms for retrieval of two dimensional ocean wave spectra from LCS. Error bars for the LCS phase, which contains information on the ocean wave phase speed and propagation direction are derived. The error estimates depend on the respective LCS coherence and the amount of smoothing applied in the LCS estimation process. A model for the LCS coherence is introduced. The first part of the model describes the dependence of the coherence on system parameters like spatial resolution. The second part is associated with the motion of the imaged ocean wave field. Decorrelation is shown to be caused by the coupling of dispersive ocean wave components in the SAR image formation process. Forward simulations for the cross spectrum coherence are carried out using parameterized swell and wind sea spectra. Coherence is estimated from a global data set of reprocessed ERS-2 wave mode data and compared to theory.

Marine parameters from synergy of optical and radar satellite data

Abstract In 2001 the European Space Agency ESA will launch the earth observation satellite ENVISAT. It will carry several instruments that provide new opportunities to measure oceanographic variables. Together, they represent the main measurement techniques of satellite oceanography, and complement each other in an ideal manner. These instruments are to be used in synergy to: • Improve the analysis of measured wind and ocean wave fields, and thereby improve weather forecasting at weather centers; • Determine the extent and variables of sea ice and develop a five-day sea ice prediction model, to support maritime shipping and offshore activities; • Monitor and map sediment and suspended matter transport in coastal regions, especially in areas with large river estuaries, which greatly affects shipping lanes, harbors, and dredging activities; • Monitor hydrobiological and bio-geochemical variables related to water quality in coastal regions and large inland waters, which affects ecology, coastal development, aquaculture, drinking water supplies, and tourism. To prepare the oceanographic community to make best use of the ENVISAT sensors in the pre-launch phase, existing algorithms to derive marine parameters are used and validated using data from the ERS SAR, the ERS RA, SeaWiFS and IRS MOS sensors now in operation. Derived products are used to address problems that can best be tackled using the synergy of radar and optical data, such as the effect of surface slicks on radar wind measurements, of sea state on ocean color, of wind and waves on the resuspension of suspended matter, and of wind and waves on sea ice variables.

Comparative study of change detection for reconstruction monitoring based on very high resolution optical data

With the rapid development of remote sensing techniques, change detection plays an important role in urban monitoring. In this paper, difference-based methods are compared and evaluated for reconstruction monitoring of urban area using very high resolution (VHR) optical data, which are classical image differencing, iteratively reweighted multivariate alteration detection (IR-MAD) and IR-MAD incorporating three kinds of textural features. The experimental results show that IR-MAD method incorporating appropriate textural feature can improve IR-MAD to some extent and image differencing has the weakest performance. Compared with manually acquired reference data, the accuracy of the change detection map extracted by IR-MAD combined with textural information is satisfying on the whole.