scholarly journals Proof of Concept for Sea Ice Stage of Development Classification Using Deep Learning

2020 ◽  
Vol 12 (15) ◽  
pp. 2486
Author(s):  
Ryan Kruk ◽  
M. Christopher Fuller ◽  
Alexander S. Komarov ◽  
Dustin Isleifson ◽  
Ian Jeffrey
Keyword(s):  
Neural Network ◽  
Sea Ice ◽  
Classification Accuracy ◽  
Primary Source ◽  
The Arctic ◽  
Proof Of Concept ◽  
Stage Of Development ◽  
Ice Concentration ◽  
Sar Data ◽  
Sar Imagery

Accurate maps of ice concentration and ice type are needed to address increased interest in commercial marine transportation through the Arctic. RADARSAT-2 SAR imagery is the primary source of data used by expert ice analysts at the Canadian Ice Service (CIS) to produce sea ice maps over the Canadian territory. This study serves as a proof of concept that neural networks can be used to accurately predict ice type from SAR data. Datasets of SAR images served as inputs, and CIS ice charts served as labelled outputs to train a neural network to classify sea ice type. Our results show that DenseNet achieves the highest overall classification accuracy of 94.0% including water and the highest ice classification accuracy of 91.8% on a three class dataset using a fusion of HH and HV SAR polarizations for the input samples. The 91.8% ice classification accuracy validates the premise that a neural network can be used to effectively categorize different ice types based on SAR data.

Convolutional Neural Networks for Classification of Sea Ice Types in Sentinel-1 SAR Data

2021 ◽  
Author(s):  
Anton Korosov ◽  
Hugo Boulze ◽  
Julien Brajard
Keyword(s):  
Random Forest ◽  
Sea Ice ◽  
Classification Accuracy ◽  
Thermal Noise ◽  
Noise Removal ◽  
The Arctic ◽  
First Year ◽  
Sar Data ◽  
The Impact

<p>A new algorithm for classification of sea ice types on Sentinel-1 Synthetic Aperture Radar (SAR) data using a convolutional neural network (CNN) is presented.  The CNN is trained on reference ice charts produced by human experts and compared with an existing machine learning algorithm based on texture features and random forest classifier. The CNN is trained on a dataset from winter 2020 for retrieval of four classes: ice free, young ice, first-year ice and old ice. The accuracy of our classification is 91.6%. The error is a bit higher for young ice (76%) and first-year ice (84%). Our algorithm outperforms the existing random forest product for each ice type. It has also proved to be more efficient in computing time and less sensitive to the noise in SAR data.</p><p> </p><p>Our study demonstrates that CNN can be successfully applied for classification of sea ice types in SAR data. The algorithm is applied in small sub-images extracted from a SAR image after preprocessing including thermal noise removal. Validation shows that the errors are mostly attributed to coarse resolution of ice charts or misclassification of training data by human experts.</p><p> </p><p>Several sensitivity experiments were conducted for testing the impact of CNN architecture, hyperparameters, training parameters and data preprocessing on accuracy. It was shown that a CNN with three convolutional layers, two max-pool layers and three hidden dense layers can be applied to a sub-image with size 50 x 50 pixels for achieving the best results. It was also shown that a CNN can be applied to SAR data without thermal noise removal on the preprocessing step. Understandably, the classification accuracy decreases to 89% but remains reasonable.</p><p> </p><p>The main advantages of the new algorithm are the ability to classify several ice types, higher classification accuracy for each ice type and higher speed of processing than in the previous studies. The relative simplicity of the algorithm (both texture analysis and classification are performed by CNN) is also a benefit. In addition to providing ice type labels, the algorithm also derives the probability of belonging to a class. Uncertainty of the method can be derived from these probabilities and used in the assimilation of ice type in numerical models. </p><p><br>Given the high accuracy and processing speed, the CNN-based algorithm is included in the Copernicus Marine Environment Monitoring Service (CMEMS) for operational sea ice type retrieval for generating ice charts in the Arctic Ocean. It is already released as an open source software and available on Github: https://github.com/nansencenter/s1_icetype_cnn.</p>

ALOS-2/PALSAR-2 dual-polarized L-band data for sea ice parameter estimation and sea ice classification 

2021 ◽  
Author(s):  
Juha Karvonen
Keyword(s):  
Parameter Estimation ◽  
Sea Ice ◽  
Sea Ice Concentration ◽  
Sea Ice Thickness ◽  
Ice Concentration ◽  
Sea Ice Drift ◽  
L Band ◽  
Sar Data ◽  
Dual Polarized ◽  
Sar Imagery

<p>This research is related to the JAXA 6th Research Announcement for the Advanced Land<br>Observing Satellite-2 (ALOS-2) project "Improved Sea Ice Parameter Estimation with L-Band SAR (ISIPELS)".<br>In the study ALOS-2/PALSAR-2 dual-polarized Horizontal-transmit-Horizontal-receive/<br>Horizontal-transmit-Vertical-receive (HH/HV) ScanSAR mode L-band  Synthetic Aperture Radar (SAR) imagery<br>over an Arctic study area were evaluated for their suitability for operational sea ice monitoring.<br>The SAR data consisting of about 140 HH/HV ScanSAR ALOS-2/PALSAR-2 images were acquired during the winter 2017.<br>These L-band SAR data were studied for estimation of different sea ice parameters:<br>sea ice concentration, sea ice thickness, sea ice type, sea ice drift. Also some comparisons with nearly<br>coincident C-band data over the same study area have been made. The results indicate that L-band<br>SAR data from ALOS-2/PALSAR-2 are very useful for estimating the studied sea ice parameters and equally good<br>or better than using the conventional operational dual-polarized C-band SAR satellite data.</p><p> </p>

scholarly journals An Optimal Decision-Tree Design Strategy and Its Application to Sea Ice Classification from SAR Imagery

2019 ◽  
Vol 11 (13) ◽  
pp. 1574 ◽  
Author(s):  
Johannes Lohse ◽  
Anthony P. Doulgeris ◽  
Wolfgang Dierking
Keyword(s):  
Decision Tree ◽  
Sea Ice ◽  
Classification Accuracy ◽  
Optimal Algorithm ◽  
Computation Time ◽  
Optimal Decision ◽  
Single Class ◽  
Feature Sets ◽  
Sar Data ◽  
Sar Imagery

We introduce the fully automatic design of a numerically optimized decision-tree algorithm and demonstrate its application to sea ice classification from SAR data. In the decision tree, an initial multi-class classification problem is split up into a sequence of binary problems. Each branch of the tree separates one single class from all other remaining classes, using a class-specific selected feature set. We optimize the order of classification steps and the feature sets by combining classification accuracy and sequential search algorithms, looping over all remaining features in each branch. The proposed strategy can be adapted to different types of classifiers and measures for the class separability. In this study, we use a Bayesian classifier with non-parametric kernel density estimation of the probability density functions. We test our algorithm on simulated data as well as airborne and spaceborne SAR data over sea ice. For the simulated cases, average per-class classification accuracy is improved between 0.5% and 4% compared to traditional all-at-once classification. Classification accuracy for the airborne and spaceborne SAR datasets was improved by 2.5% and 1%, respectively. In all cases, individual classes can show larger improvements up to 8%. Furthermore, the selection of individual feature sets for each single class can provide additional insights into physical interpretation of different features. The improvement in classification results comes at the cost of longer computation time, in particular during the design and training stage. The final choice of the optimal algorithm therefore depends on time constraints and application purpose.

scholarly journals Retrieval of Snow Depth over Arctic Sea Ice Using a Deep Neural Network

2019 ◽  
Vol 11 (23) ◽  
pp. 2864 ◽  
Author(s):  
Jiping Liu ◽  
Yuanyuan Zhang ◽  
Xiao Cheng ◽  
Yongyun Hu
Keyword(s):  
Neural Network ◽  
Sea Ice ◽  
Snow Depth ◽  
Deep Neural Network ◽  
Arctic Basin ◽  
Temporal Variations ◽  
Arctic Sea Ice ◽  
The Arctic ◽  
Sea Ice Thickness ◽  
Arctic Sea

The accurate knowledge of spatial and temporal variations of snow depth over sea ice in the Arctic basin is important for understanding the Arctic energy budget and retrieving sea ice thickness from satellite altimetry. In this study, we develop and validate a new method for retrieving snow depth over Arctic sea ice from brightness temperatures at different frequencies measured by passive microwave radiometers. We construct an ensemble-based deep neural network and use snow depth measured by sea ice mass balance buoys to train the network. First, the accuracy of the retrieved snow depth is validated with observations. The results show the derived snow depth is in good agreement with the observations, in terms of correlation, bias, root mean square error, and probability distribution. Our ensemble-based deep neural network can be used to extend the snow depth retrieval from first-year sea ice (FYI) to multi-year sea ice (MYI), as well as during the melting period. Second, the consistency and discrepancy of snow depth in the Arctic basin between our retrieval using the ensemble-based deep neural network and two other available retrievals using the empirical regression are examined. The results suggest that our snow depth retrieval outperforms these data sets.

scholarly journals Sea Ice Concentration Products over Polar Regions with Chinese FY3C/MWRI Data

2021 ◽  
Vol 13 (11) ◽  
pp. 2174
Author(s):  
Lijian Shi ◽  
Sen Liu ◽  
Yingni Shi ◽  
Xue Ao ◽  
Bin Zou ◽  
...  
Keyword(s):  
Time Series ◽  
Sea Ice ◽  
Ocean Circulation ◽  
The Arctic ◽  
Retrieval Algorithm ◽  
Atmospheric Effects ◽  
Observation Data ◽  
Long Time Series ◽  
Ice Concentration ◽  
Long Time

Polar sea ice affects atmospheric and ocean circulation and plays an important role in global climate change. Long time series sea ice concentrations (SIC) are an important parameter for climate research. This study presents an SIC retrieval algorithm based on brightness temperature (Tb) data from the FY3C Microwave Radiation Imager (MWRI) over the polar region. With the Tb data of Special Sensor Microwave Imager/Sounder (SSMIS) as a reference, monthly calibration models were established based on time–space matching and linear regression. After calibration, the correlation between the Tb of F17/SSMIS and FY3C/MWRI at different channels was improved. Then, SIC products over the Arctic and Antarctic in 2016–2019 were retrieved with the NASA team (NT) method. Atmospheric effects were reduced using two weather filters and a sea ice mask. A minimum ice concentration array used in the procedure reduced the land-to-ocean spillover effect. Compared with the SIC product of National Snow and Ice Data Center (NSIDC), the average relative difference of sea ice extent of the Arctic and Antarctic was found to be acceptable, with values of −0.27 ± 1.85 and 0.53 ± 1.50, respectively. To decrease the SIC error with fixed tie points (FTPs), the SIC was retrieved by the NT method with dynamic tie points (DTPs) based on the original Tb of FY3C/MWRI. The different SIC products were evaluated with ship observation data, synthetic aperture radar (SAR) sea ice cover products, and the Round Robin Data Package (RRDP). In comparison with the ship observation data, the SIC bias of FY3C with DTP is 4% and is much better than that of FY3C with FTP (9%). Evaluation results with SAR SIC data and closed ice data from RRDP show a similar trend between FY3C SIC with FTPs and FY3C SIC with DTPs. Using DTPs to present the Tb seasonal change of different types of sea ice improved the SIC accuracy, especially for the sea ice melting season. This study lays a foundation for the release of long time series operational SIC products with Chinese FY3 series satellites.

scholarly journals Sea Ice Concentration and Sea Ice Extent Mapping with L-Band Microwave Radiometry and GNSS-R Data from the FFSCat Mission Using Neural Networks

2021 ◽  
Vol 13 (6) ◽  
pp. 1139
Author(s):  
David Llaveria ◽  
Juan Francesc Munoz-Martin ◽  
Christoph Herbert ◽  
Miriam Pablos ◽  
Hyuk Park ◽  
...  
Keyword(s):  
Sea Ice ◽  
Microwave Radiometer ◽  
Cost Effective ◽  
The Arctic ◽  
Sea Ice Concentration ◽  
Sea Ice Extent ◽  
Neural Network Approach ◽  
Ice Concentration ◽  
L Band ◽  
Moderate Cost

CubeSat-based Earth Observation missions have emerged in recent times, achieving scientifically valuable data at a moderate cost. FSSCat is a two 6U CubeSats mission, winner of the ESA S3 challenge and overall winner of the 2017 Copernicus Masters Competition, that was launched in September 2020. The first satellite, 3Cat-5/A, carries the FMPL-2 instrument, an L-band microwave radiometer and a GNSS-Reflectometer. This work presents a neural network approach for retrieving sea ice concentration and sea ice extent maps on the Arctic and the Antarctic oceans using FMPL-2 data. The results from the first months of operations are presented and analyzed, and the quality of the retrieved maps is assessed by comparing them with other existing sea ice concentration maps. As compared to OSI SAF products, the overall accuracy for the sea ice extent maps is greater than 97% using MWR data, and up to 99% when using combined GNSS-R and MWR data. In the case of Sea ice concentration, the absolute errors are lower than 5%, with MWR and lower than 3% combining it with the GNSS-R. The total extent area computed using this methodology is close, with 2.5% difference, to those computed by other well consolidated algorithms, such as OSI SAF or NSIDC. The approach presented for estimating sea ice extent and concentration maps is a cost-effective alternative, and using a constellation of CubeSats, it can be further improved.

scholarly journals Retrieval of Summer Sea Ice Concentration in the Pacific Arctic Ocean from AMSR2 Observations and Numerical Weather Data Using Random Forest Regression

2021 ◽  
Vol 13 (12) ◽  
pp. 2283
Author(s):  
Hyangsun Han ◽  
Sungjae Lee ◽  
Hyun-Cheol Kim ◽  
Miae Kim
Keyword(s):  
Random Forest ◽  
Sea Ice ◽  
Arctic Ocean ◽  
Open Water ◽  
The Arctic ◽  
Atmospheric Effects ◽  
Sea Ice Concentration ◽  
Air Temperatures ◽  
The Pacific ◽  
Ice Concentration

The Arctic sea ice concentration (SIC) in summer is a key indicator of global climate change and important information for the development of a more economically valuable Northern Sea Route. Passive microwave (PM) sensors have provided information on the SIC since the 1970s by observing the brightness temperature (TB) of sea ice and open water. However, the SIC in the Arctic estimated by operational algorithms for PM observations is very inaccurate in summer because the TB values of sea ice and open water become similar due to atmospheric effects. In this study, we developed a summer SIC retrieval model for the Pacific Arctic Ocean using Advanced Microwave Scanning Radiometer 2 (AMSR2) observations and European Reanalysis Agency-5 (ERA-5) reanalysis fields based on Random Forest (RF) regression. SIC values computed from the ice/water maps generated from the Korean Multi-purpose Satellite-5 synthetic aperture radar images from July to September in 2015–2017 were used as a reference dataset. A total of 24 features including the TB values of AMSR2 channels, the ratios of TB values (the polarization ratio and the spectral gradient ratio (GR)), total columnar water vapor (TCWV), wind speed, air temperature at 2 m and 925 hPa, and the 30-day average of the air temperatures from the ERA-5 were used as the input variables for the RF model. The RF model showed greatly superior performance in retrieving summer SIC values in the Pacific Arctic Ocean to the Bootstrap (BT) and Arctic Radiation and Turbulence Interaction STudy (ARTIST) Sea Ice (ASI) algorithms under various atmospheric conditions. The root mean square error (RMSE) of the RF SIC values was 7.89% compared to the reference SIC values. The BT and ASI SIC values had three times greater values of RMSE (20.19% and 21.39%, respectively) than the RF SIC values. The air temperatures at 2 m and 925 hPa and their 30-day averages, which indicate the ice surface melting conditions, as well as the GR using the vertically polarized channels at 23 GHz and 18 GHz (GR(23V18V)), TCWV, and GR(36V18V), which accounts for atmospheric water content, were identified as the variables that contributed greatly to the RF model. These important variables allowed the RF model to retrieve unbiased and accurate SIC values by taking into account the changes in TB values of sea ice and open water caused by atmospheric effects.

scholarly journals Deep Learning Convolutional Neural Network Applying for the Arctic Acoustic Tomography Current Inversion Accuracy Improvement

2021 ◽  
Vol 9 (7) ◽  
pp. 755
Author(s):  
Kangkang Jin ◽  
Jian Xu ◽  
Zichen Wang ◽  
Can Lu ◽  
Long Fan ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Sea Ice ◽  
Convolutional Neural Network ◽  
Arctic Sea Ice ◽  
The Arctic ◽  
Small Scale ◽  
Acoustic Tomography ◽  
Arctic Sea ◽  
Current Inversion

Warm current has a strong impact on the melting of sea ice, so clarifying the current features plays a very important role in the Arctic sea ice coverage forecasting study field. Currently, Arctic acoustic tomography is the only feasible method for the large-range current measurement under the Arctic sea ice. Furthermore, affected by the high latitudes Coriolis force, small-scale variability greatly affects the accuracy of Arctic acoustic tomography. However, small-scale variability could not be measured by empirical parameters and resolved by Regularized Least Squares (RLS) in the inverse problem of Arctic acoustic tomography. In this paper, the convolutional neural network (CNN) is proposed to enhance the prediction accuracy in the Arctic, and especially, Gaussian noise is added to reflect the disturbance of the Arctic environment. First, we use the finite element method to build the background ocean model. Then, the deep learning CNN method constructs the non-linear mapping relationship between the acoustic data and the corresponding flow velocity. Finally, the simulation result shows that the deep learning convolutional neural network method being applied to Arctic acoustic tomography could achieve 45.87% accurate improvement than the common RLS method in the current inversion.

Sign in / Sign up

Export Citation Format

Share Document