Virtual radar ice buoys  –  a method for measuring fine-scale dynamic properties of sea ice

Abstract. Here we present an algorithm for continuous ice dynamics estimation based on coastal and ship radar data. The ice dynamics are estimated based on automatically selected ice targets in the images. These targets are here called virtual buoys (VB's) and are tracked based on an optical flow method. To maintain continuous ice drift tracking new VB's are added after a~given number of VB's have been lost i.e. they can not be tracked reliably any more. Some tracking results and some computed derived quantities for a~few test cases are presented.

Download Full-text

Virtual radar ice buoys – a method for measuring fine-scale sea ice drift

The Cryosphere ◽

10.5194/tc-10-29-2016 ◽

2016 ◽

Vol 10 (1) ◽

pp. 29-42 ◽

Cited By ~ 13

Author(s):

J. Karvonen

Keyword(s):

Sea Ice ◽

Dynamic Properties ◽

Weather Conditions ◽

Radar Data ◽

Optical Flow Method ◽

Drift Estimation ◽

Ice Drift ◽

Sea Ice Drift ◽

Ice Field ◽

Direction Information

Abstract. Here we present an algorithm for continuous ice drift estimation based on coastal and ship radar data. The ice drift is estimated for automatically selected ice targets in the images. These targets are here called virtual buoys (VBs) and are tracked based on an optical flow method. To maintain continuous ice drift tracking new VBs are added after a given number of VBs have been lost; i.e. they can not be tracked reliably any more. Here we also apply the algorithm to data of three test cases to demonstrate its capabilities and properties. Two of these cases use coastal radar data and one ship radar data. Ice drift velocity and direction information derived from the VB motion are computed and compared to the prevailing ice and weather conditions. Also a quantity measuring the local divergence or convergence is computed for some VBs to demonstrate the capability to estimate derived kinematic sea ice parameters from VB location time series. The results produced by the algorithm can be used as an input for estimation of the dynamic properties of sea the ice field, such as ice divergence or convergence, shear, vorticity, and total deformation.

Download Full-text

A Precipitation Nowcasting Mechanism for Real-World Data Based on Machine Learning

Mathematical Problems in Engineering ◽

10.1155/2020/8408931 ◽

2020 ◽

Vol 2020 ◽

pp. 1-11

Author(s):

Yanfei Xiang ◽

Jianbing Ma ◽

Xi Wu

Keyword(s):

Machine Learning ◽

Optical Flow ◽

Weather Prediction ◽

Radar Data ◽

Machine Learning Techniques ◽

Economic Losses ◽

Model Parameters ◽

Real World Data ◽

Flow Method ◽

Optical Flow Method

Unpredicted precipitations, even mild, may cause severe economic losses to many businesses. Precipitation nowcasting is hence significant for people to make correct decisions timely. For traditional methods, such as numerical weather prediction (NWP), the accuracy is limited because the smaller scale of strong convective weather must be smaller than the minimum scale that the model can capture. And it often requires a supercomputer. Furthermore, the optical flow method has been proved to be available for precipitation nowcasting. However, it is difficult to determine the model parameters because the two steps of tracking and extrapolation are separate. In contrast, current machine learning applications are based on well-selected full datasets, ignoring the fact that real datasets quite often contain missing data requiring extra consideration. In this paper, we used a real Hubei dataset in which a few radar echo data are missing and proposed a proper mechanism to deal with the situation. Furthermore, we proposed a novel mechanism for radar reflectivity data with single altitudes or cumulative altitudes using machine learning techniques. From the experimental results, we conclude that our method can predict future precipitation with a high accuracy when a few data are missing, and it outperforms the traditional optical flow method. In addition, our model can be used for various types of radar data with a type-specific feature extraction, which makes the method more flexible and suitable for most situations.

Download Full-text