Hyangsun Han

Landfast sea ice monitoring using multisensor fusion in the Antarctic

Landfast sea ice (fast ice) means sea ice that is attached to the shoreline with little or no motion in contrast to pack ice which drifts on the sea. As fast ice plays an important role in the environmental and biological systems of the Antarctic, it is crucial to accurately monitor the spatiotemporal distribution of fast ice. Previous studies on fast ice using satellite remote sensing were mostly focused on the Arctic and near-Arctic areas, whereas few studies were conducted over the Antarctic, especially the West Antarctic region. This research mapped fast ice using multisensor data from 2003 to 2008 based on machine learning approaches – decision trees (DTs) and random forest (RF). A total of seven satellite-derived products, including Advanced Microwave Scanning Radiometer for the Earth observing system brightness temperatures and sea ice concentration, Moderate Resolution Imaging Spectroradiometer (MODIS) ice surface temperature (IST) and Special Sensor Microwave/Imager ice velocity, were used as input variables for identifying fast ice. RF resulted in better performance than that of DT for fast ice classification. Visual comparison of the fast ice classification results with 250-m MODIS images for selected areas also revealed that RF outperformed DT. Ice velocity and IST were identified as the most contributing variables to classify fast ice. Spatiotemporal variations of fast ice in the East and West Antarctic were also examined using the time series of the fast ice maps produced by RF. The residence time of fast ice was much shorter in the West Antarctic than in the East.

Detection of Convective Initiation Using Meteorological Imager Onboard Communication, Ocean, and Meteorological Satellite Based on Machine Learning Approaches

As convective clouds in Northeast Asia are accompanied by various hazards related with heavy rainfall and thunderstorms, it is very important to detect convective initiation (CI) in the region in order to mitigate damage by such hazards. In this study, a novel approach for CI detection using images from Meteorological Imager (MI), a payload of the Communication, Ocean, and Meteorological Satellite (COMS), was developed by improving the criteria of the interest fields of Rapidly Developing Cumulus Areas (RDCA) derivation algorithm, an official CI detection algorithm for Multi-functional Transport SATellite-2 (MTSAT-2), based on three machine learning approaches—decision trees (DT), random forest (RF), and support vector machines (SVM). CI was defined as clouds within a 16 × 16 km window with the first detection of lightning occurrence at the center. A total of nine interest fields derived from visible, water vapor, and two thermal infrared images of MI obtained 15–75 min before the lightning occurrence were used as input variables for CI detection. RF produced slightly higher performance (probability of detection (POD) of 75.5% and false alarm rate (FAR) of 46.2%) than DT (POD of 70.7% and FAR of 46.6%) for detection of CI caused by migrating frontal cyclones and unstable atmosphere. SVM resulted in relatively poor performance with very high FAR ~83.3%. The averaged lead times of CI detection based on the DT and RF models were 36.8 and 37.7 min, respectively. This implies that CI over Northeast Asia can be forecasted ~30–45 min in advance using COMS MI data.

Evaluation of SSM/I and AMSR-E Sea Ice Concentrations in the Antarctic Spring Using KOMPSAT-1 EOC Images

To evaluate sea ice concentrations (SICs) from the special sensor microwave/imager (SSM/I) and advanced microwave scanning radiometer-EOS (AMSR-E), we observed sea ice with the 6-m-resolution panchromatic electronic optical camera (EOC) sensor onboard the Korea Multi-Purpose Satellite-1 (KOMPSAT-1). A total of 68 cloud-free EOC images were obtained across the Antarctic continental edges from September to November 2005. Sea ice types in the EOC images were classified into white ice (W), gray ice (G), and dark-gray ice (D) and then compared with SSM/I and AMSR-E SICs. Spatiotemporal standard deviation of passive microwave SIC proved useful in selecting temporally stable and spatially homogeneous SICs to overcome the diurnal variation of sea ice in the analysis of data from multiple satellites. In the Antarctic spring, the EOC SIC of W + G showed the best fit to SSM/I SIC calculated by the NASA Team (NT) algorithm (mean difference of -2.3% and rmse of 3.2%), whereas that of W + G + D showed the best fit to AMSR-E SIC calculated by the NT2 algorithm (mean difference of 0.3% and rmse of 1.4%). It is concluded that the SSM/I NT algorithm responds to young ice in addition to the ice types A and B, whereas the AMSR-E NT2 algorithm detects ice type C and thin ice as well. The 4.7% difference of SICs between AMSR-E and SSM/I was attributed to the enhanced detection of ice type C (2.1%) and thin ice (2.6%) of the AMSR-E NT2 algorithm.

Retrieval of Melt Ponds on Arctic Multiyear Sea Ice in Summer from TerraSAR-X Dual-Polarization Data Using Machine Learning Approaches: A Case Study in the Chukchi Sea with Mid-Incidence Angle Data

Melt ponds, a common feature on Arctic sea ice, absorb most of the incoming solar radiation and have a large effect on the melting rate of sea ice, which significantly influences climate change. Therefore, it is very important to monitor melt ponds in order to better understand the sea ice-climate interaction. In this study, melt pond retrieval models were developed using the TerraSAR-X dual-polarization synthetic aperture radar (SAR) data with mid-incidence angle obtained in a summer multiyear sea ice area in the Chukchi Sea, the Western Arctic, based on two rule-based machine learning approaches—decision trees (DT) and random forest (RF)—in order to derive melt pond statistics at high spatial resolution and to identify key polarimetric parameters for melt pond detection. Melt ponds, sea ice and open water were delineated from the airborne SAR images (0.3-m resolution), which were used as a reference dataset. A total of eight polarimetric parameters (HH and VV backscattering coefficients, co-polarization ratio, co-polarization phase difference, co-polarization correlation coefficient, alpha angle, entropy and anisotropy) were derived from the TerraSAR-X dual-polarization data and then used as input variables for the machine learning models. The DT and RF models could not effectively discriminate melt ponds from open water when using only the polarimetric parameters. This is because melt ponds showed similar polarimetric signatures to open water. The average and standard deviation of the polarimetric parameters based on a 15 × 15 pixel window were supplemented to the input variables in order to consider the difference between the spatial texture of melt ponds and open water. Both the DT and RF models using the polarimetric parameters and their texture features produced improved performance for the retrieval of melt ponds, and RF was superior to DT. The HH backscattering coefficient was identified as the variable contributing the most, and its spatial standard deviation was the next most contributing one to the classification of open water, sea ice and melt ponds in the RF model. The average of the co-polarization phase difference and the alpha angle in a mid-incidence angle were also identified as the important variables in the RF model. The melt pond fraction and sea ice concentration retrieved from the RF-derived melt pond map showed root mean square deviations of 2.4% and 4.9%, respectively, compared to those from the reference melt pond maps. This indicates that there is potential to accurately monitor melt ponds on multiyear sea ice in the summer season at a local scale using high-resolution dual-polarization SAR data.

Construction and application of tomographic SAR system based on GB-SAR system

Synthetic Aperture Radar (SAR) provides high-resolution images regardless of weather conditions and solar illumination. However, a SAR image projects the 3D distributed targets onto a 2D plane in range and azimuth. Tomographic SAR (TomoSAR) is a technique to generate 3D image by adding the vertical baselines to the conventional SAR image acquisition. TomoSAR has recently been used to map the internal structure of forests and the geometry of the buildings by using airborne and ground-based SAR systems. In this paper, we report the construction of a Ground-Based TomoSAR (GB-TomoSAR) system and explain the basic principle of the GB-TomoSAR focusing algorithm. A GB-TomoSAR experiment was performed on the rooftop of a building at Kangwon National University. Heights of various targets were measured by generating the slice images both in the horizontal and vertical direction, which well matched to the in situ data. The experimental image has limitations in the detection of trees due to temporal decorrelation during the 8 hours of acquisition for a single tomogram. In order to overcome this limitation, we are planning to enhance the scanning speed of GB-TomoSAR in the near future.

Surface strain rates and crevassing of Campbell Glacier Tongue in East Antarctica analysed by tide-corrected DInSAR

ABSTRACT This research presents the measurement of surface strain rates over Campbell Glacier Tongue (CGT) in East Antarctica from tide-corrected ice velocity derived by removing the vertical tidal deflection from the 14 COSMO-SkyMed 1-day differential interferometric synthetic aperture radar images obtained in 2011. The hinge zone of CGT shows large longitudinal, transverse, shear, and vertical strain rates, especially in a heavy crevassing band, while the freely floating zone shows very small strain rates. This represents that the surface deformation by ice flow in the hinge zone is significantly larger than that in the freely floating zone. The orientations of crevasses in the hinge zone agree reasonably well with the directions orthogonal to the axes of the most tensile strain rates calculated from the tide-corrected ice velocity, which demonstrates that the crevassing is attributed to the gravitational ice flow rather than to the tidal flexure.

COSMO-SkyMed AO projects - Tidal deflection characteristics of Campbell Glacier, East Antarctica, observed by double differential SAR interferometry

In this paper, we suggest a method of estimating flexural parameters of glacier responding to tide height. During 2010 to 2011, we obtained 16 COSMO-SkyMed one-day interferometric tandem pairs over Campbell Glacier, East Antarctica, in every 16 days and extracted signal from vertical tidal motion by applying Double-Differential Interferometric SAR (DDInSAR) technique. To find the tidal flexural parameters for Campbell Glacier, we used several tide models (TPXO7.1, FES2004, CATS2008a and Ross_Inv) and an elastic beam model. The result showed that the inverse barometric effect-corrected Ross_Inv showed best correlation with the tidal motion from DDInSAR. The deflection constant of Campbell Glacier was determined to be 0.926 with R2 of 0.964. The elastic damping factor of a stream line, as an example, was derived as 0.827±0.099 km-1, from which the glacier thickness was estimated to be 406±52 m.

Comparison of SSM/I Sea Ice Concentration with Kompsat-1 EOC Images of the Arctic and Antarctic

We have compared Kompsat-1 EOC images (6.6 m resolution, panchromatic) of the Arctic and Antarctic sea ice with SSM/I sea ice concentration (SIC). EOC images were obtained from 10 orbits (624 scenes) across the Arctic sea ice edges from July to August and 11 orbits (676 scenes) across the Antarctic continental edges from September to November, all in year 2005. By applying supervised classification and visual identification to about 12% usable images out of the total scenes, we have classified various sea ice types such as multi-year ice and first- year ice (M+F), Young ice (Y), and New ice (N). EOC SIC were derived and compared with SSMI SIC calculated by NASA Team Algorithm (NTA). In summertime of the Arctic, the correlation coefficient between EOC SIC (M+F+Y+N) and SSM/I SIC were low (0.671) due to rapid temporal and spatial variation of sea ice. For springtime in the Antarctic, EOC SIC (M+F+Y, excluding N) and SSM/I SIC have shown the highest correlation coefficient (0.873). We have concluded that the NTA- derived SSM/I SIC includes Y as well as M+F, but not N.

A study of the feasibility of using KOMPSAT-5 SAR data to map sea ice in the Chukchi Sea in late summer

ABSTRACT In this study, a sea ice mapping model based on Random Forest (RF), a rule-based machine learning approach, has been developed for the Korea Multi-Purpose Satellite-5 (KOMPSAT-5) Synthetic Aperture Radar (SAR) data in Enhanced Wide swath mode obtained from 6 August to 9 September 2015 in the Chukchi Sea. A total of 12 texture features derived from backscattering intensity and the gray-level co-occurrence matrix were used as input variables for sea ice mapping. The RF model produced a sea ice map with a grid spacing of 125 m, demonstrating excellent performance in the classification of sea ice and open water with an overall accuracy of 99.2% and a kappa coefficient of 98.5%. Sea ice concentration (SIC) retrieved from the RF-derived sea ice maps was compared with that from ice charts. The mean and median values of the differences between the SICs derived from the RF model and the ice charts were −8.85% and −8.38%, respectively. Such difference was attributed to both the uncertainty in the ice charts and classification error of the RF model.

Radar Backscattering of Lake Ice During Freezing and Thawing Stages Estimated by Ground-Based Scatterometer Experiment and Inversion From Genetic Algorithm

Lake ice under phase transition shows large variation on radar backscattering due to the changes of dielectric constant and roughness of ice surface and thus the transmissivity of microwave into ice body. To study the effects of freezing/thawing of ice on radar backscattering in a short time, we spread water over lake ice and continuously measured radar backscattering by using a ground-based microwave scatterometer system operated in C-band HH polarization. By establishing scattering models and applying inversion from genetic algorithm, radar returns were separated into ice-surface, volume, and ice-bottom scatterings, and the changes in dielectric constant and roughness parameters of ice surface were estimated as well. Immediately after spreading water on ice surface, ice-surface scattering was strongest due to high dielectric constant of surface water while volume and ice-bottom scatterings were very weak due to low microwave transmissivity into ice body. As surface water was being frozen, ice-surface scattering became weak with decreasing dielectric constant while volume and ice-bottom scattering increased due to higher transmissivity into ice body. In a transition stage, when surface water was almost frozen, all three scatterings increased simultaneously. Crystallization of ice produced rougher surface overcoming the decrease in dielectric constant, resulting in the increase of ice-surface scattering, while volume and ice-bottom scattering was continuously increased due to increasing transmissivity. At the end of the experiment, air temperature rose above freezing point, and ice surface thawed again so that ice-surface scattering increased while volume and ice-bottom scattering were decreased.