Search

Compact-polarimetry (CP) synthetic aperture radar (SAR) observations are presented for major sea ice types in each ice season. CP data for three wide-swath Radarsat Constellation Mission (RCM) modes were simulated and evaluated. Regression models and statistical distances as functions of incidence angles were calculated for 26 CP parameters, based on 969 samples of user-selected homogeneous regions of sea ice. CP parameters, best able to discriminate sea ice types and open water, were quantitatively identified in three incidence angle ranges (19–29°, 30–39°, 40–49°). These parameters will likely provide discrimination of sea ice types and open water for both visual interpretation and automated classification. Several parameter–ice type combinations exhibit novel scattering responses, which present new opportunities for ice type discrimination and for inferring scattering mechanisms. Specifically, phase-related parameters with early-stage ice types provide discrimination for ice type pairings tha...

The mapping and monitoring of sea ice regions represents a key application area for spaceborne synthetic aperture radar (SAR) missions. The European Space Agency (ESA) is under-taking the development of Sentinel-1, an imaging radar mission at C-Band for the continuation of SAR operational applications. Monitoring sea ice zones and the arctic environment is one of the major application areas supported by Sentinel-1. Other spaceborne SAR missions are being developed or considered by a number of space organizations. The study described herein focuses on evaluating the contribution of SAR constellations to operational and scientific monitoring of sea ice conditions through the analysis of multi-parameter SAR datasets (airborne, spaceborne) and through the development and exploitation of a simulation tool able to predict sea ice radar signatures as a function of the ice type and condition and of sensor parameters.

A sea ice classification technique is introduced. The classification was performed on Radarsat-2 dual polarization image (07.02.2009) and MODIS images (09.02.2009) that covered coastal area in the western part of Baltic Sea. The classification on SAR image was based on Shannon entropy differences that occur on various ice types. Analysis showed that four ice types (water, fast ice, ridged ice and deformed ice) were identifiable. Also two independent analysis were performed on MODIS image. The spectral classification enabled to separate ridged ice regions and fast ice areas. Using histogram analysis ice concentration map was created. The analysis of Modis image enabled to identify regions where water was covering ice surface. The same phenomena was also detected from SAR data.

Fully automatic algorithms have been developed which can produce sea ice type classification maps and sea ice concentration estimates of synthetic aperture radar (SAR) imagery. The sea ice type classification algorithm uses local statistics to determine ice type boundaries, and the ice concentration algorithm iteratively decomposes the histogram into ice and water histograms. algorithms have been used on both simulated imagery and actual SAR imagery gathered during the 1984 and

In April-May 1998 sea ice conditions in the Barents and Kara Seas were studied from synchronous RADARSAT, ERS and "Okean" radar images. SAR images were transmitted in quasi-real time onboard nuclear icebreaker "Sovetsky Soyuz" and signatures of the main sea ice types and features were retrieved and compared with subsatellite observations. RADARSAT and ERS signatures of first-year ice are rather similar, whereas for gray and new ice they can be different. Analysis of synchronous RADARSAT and "Okean" images proved, that large-scale ice features can be identified from both images, whereas navigationally important small-scale features-only from SAR.

Coincident ship based X-, C-, and L-band scatterometer data and measurements of physical ice and snow properties were collected during ERS-1 synthetic aperture radar (SAR) overflights in the Barents Sea in March of 1992. One aspect of this Seasonal Ice Zone Experiment (SIZEX) concentrated on the marginal ice zone (MIZ), where a variety of sea ice types and open ocean conditions were encountered. A specific objective of SIZEX was to determine which MIZ ice types can be clearly discerned using ERS-1 SAR C-band imagery. First- and second-order statistics were generated to characterize ERS-1 ice signature. Comparing these statistics for each ice type, reveals that many of the ice types observed in the MIZ have comparable tones, as indicated by their means (i.e., multiyear and pancakes). However, the use of the SAR image texture may be used in conjunction with the tonal properties to successfully discern the various ice types and open water (all wind regimes) found in the MIZ. >

During a field expedition with the Russian nuclear icebreaker "Sovetsky Soyuz" to the Laptev Sea in September 1997, RADARSAT ScanSAR images were used to investigate late summer ice conditions and support ice navigation. The predominant ice types were first year and newly formed ice, primarily nilas. The in situ observations made from the icebreaker and its helicopter were used to validate the ScanSAR observations and assess the capability of RADARSAT data to discriminate open water, first year and new ice. Also SSM/I and ERS SAR were compared with the ScanSAR data. To observe first year ice at concentrations above 10% seemed to be straight forward, while nilas was difficult to discriminate from open water during low wind conditions. Use of RADARSAT ScanSAR data represents a significant improvement in ice monitoring and navigation in the Northern Sea Route.

This paper describes a minor part of the work done in connection with a preliminary investigation of a neural network’s capability to classify ice types. It includes a short review of earlier used techniques, implementation of different neural networks and results from various experiments with these networks. The estimation of ice type concentrations from Synthetic Aperture Radar (SAR) images has been investigated for several years, see e.g. [9]. The classification estimation has been performed by training a Bayesian Maximum Likelihood Classifier (BMLC) [8] with a classification rate about 80%. The neural networks considered are all of the feed-forward type. For training, different learning algorithms and error functions are used. Both pruning and construction algorithms are used to get an optimal architecture. Experiments showed that almost any kind of neural network, using a Standard Back-Propagation (Std_BP) learning algorithm for minimising the Mean Square Error (MSE), is able to perform better than the BMLC. The reason is that the neural network is able to use a larger training set than the BMLC.

ERS-1 Synthetic Aperture Radar (SAR) imagery from the Alaska SAR facility is available to operational ice analysts at the Joint Ice Center. Analyst can use the imagery directly through manual interpretation. One such use would be to provide information on ice conditions for a field expedition. In addition, analysts can run automated algorithms which produce estimates of ice type and ice motion from SAR images. Such automated analysis will become increasingly important if SAR is to be used to improve climatologies of ice characteristics and to improve the performance of dynamic ice models. Here the ice classification and ice motion algorithms developed by the Jet Propulsion Laboratory for ERS-1 SAR imagery are evaluated and results presented. Remote sensing, Synthetic aperture radar, Sea ice.