Ice Prediction Research Articles

DLR related model development and performance analysis in the framework of FLEXITRANSTORE

This paper demonstrates a practical implementation of a dynamic line rating (DLR) system and an effective ice prediction and prevention method using line monitoring sensors within the FLEXITRANSTORE international project framework. The presented anti-icing system operated continuously for two years. As a part of the project, 4 line monitoring sensors and 2 weather stations were installed on that section of the power line where icing is most likely to occur. These sensors worked continuously and reliably during the demonstration period. The sensor measurements were supplemented with the load value from the SCADA system and weather forecasts, with the help of which, in addition to the real-time line rating and sag-clearance analysis, a predetermined line rating could also be determined. The investigated system also included an ice prediction subsystem as well as a sag-calculation-based ice detection subsystem. Based on the comparison of the measured and calculated parameters, it can be said that both the operation of the sensors and the quality of the data provided by them were satisfactory. In the case of conductor temperature calculation, the developed mathematical models, which can be applied independently, provided information to the system operator with almost similar accuracy than sensors’ measurements. During the observed period, the ice prediction and detection subsystem were also tested and functioned correctly. As an overview of the project, it can be concluded that an efficient DLR-sensor-based anti-icing system is possible to be realized on the high voltage distribution power grid.

A New Ice Quality Prediction Method of Wind Turbine Impeller Based on the Deep Neural Network

More and more wind turbines are installed in cold regions because of better wind resources. In these regions, the high humidity and low temperatures in winter will lead to ice accumulation on the wind turbine impeller. A different icing location or mass will lead to different natural frequency variations of the impeller. In order to monitor the icing situation in time and in advance, a method based on depth neural network technology to predict the icing mass is explored and proposed. Natural-environment icing experiments and iced-impeller modal experiments are carried out, aiming at a 600 W wind turbine, respectively. The mapping relationship between the change rate of the natural frequency of the iced impeller at different icing positions and the icing mass is obtained, and the correlation coefficients are all above 0.93. A deep neural network (DNN) prediction model of ice-coating quality for the impeller was constructed with the change rate of the first six-order natural frequencies as the input factor. The results show that the MAE and MSE of the trained model are close to 0. The average prediction error of the DNN model is 4.79%, 9.35%, 3.62%, 1.63%, respectively, under different icing states of the impeller. It can be seen that the DNN shows the best prediction ability among other methods. The smaller the actual ice-covered mass of the impeller, the larger the relative error of the ice-covered mass predicted by the DNN model. In the same ice-covered state, the relative error will decrease gradually with the increase in ice-covered mass. In a word, using the natural frequency change rate to predict the icing quality is feasible and accurate. The research achievements shown here can provide a new idea for wind farms to realize efficient and intelligent icing monitoring and prediction, provide engineering guidance for the wind turbine blade anti-icing and deicing field, and further reduce the negative impact of icing on wind power generation.

Development of black ice prediction model using GIS-based multi-sensor model validation

Abstract. Fog, freezing rain, and snow (melt) quickly condense on road surfaces, forming black ice that is difficult to identify and causes major accidents on highways. As a countermeasure to prevent icing car accidents, it is necessary to predict the amount and location of black ice. This study advanced previous models through machine learning and multi-sensor-verified results. Using spatial (hill shade, river system, bridge, and highway) and meteorological (air temperature, cloudiness, vapour pressure, wind speed, precipitation, snow cover, specific heat, latent heat, and solar radiation energy) data from the study area (Suncheon–Wanju Highway in Gurye-gun, Jeollanam-do, South Korea), the amount and location of black ice were modelled based on system dynamics to predict black ice and then simulated with a geographic information system in units of square metres. The intermediate factors calculated as input factors were road temperature and road moisture, modelled using a deep neural network (DNN) and numerical methods. Considering the results of the DNN, the root mean square error was improved by 148.6 % and reliability by 11.43 % compared to a previous study (linear regression). Based on the model results, multiple sensors were buried at four selected points in the study area. The model was compared with sensor data and verified with the upper-tailed test (with a significance level of 0.05) and fast Fourier transform (freezing does not occur when frequency = 0.00001 Hz). Results of the verified simulation can provide valuable data for government agencies like road traffic authorities to prevent traffic accidents caused by black ice.

An Internet-of-Things-Enabled System for Road Icing Detection and Prediction

Road icing has become one of the most critical factors threatening traffic safety. This article proposes an Internet of Things (IoT)-enabled road icing detection and prediction system. In the proposed system, we first design a low-power icing sensor equipped with IoT function to periodically collect current road status and transmit the sampled data to IoT gateway through Long Range Radio (LoRa). Then, we design a simple but effective algorithm deployed on IoT gateway to identify road icing in time. The algorithm is proposed based on the change trend of the sampled data of the road state, and can be adapted to the icing recognition on the road covered with various impurities. Furthermore, we put forward a newly designed deep neural network model called Trans-CGAN to achieve accurate road icing prediction even the positive and negative samples are imbalanced. Through a real system deployment and experiments, the results show that our proposed system can detect the formation of road icing effectively and timely, and shows better prediction performance of road icing than several representative models.

Establishment of icing prediction model of asphalt pavement based on support vector regression algorithm and Bayesian optimization

• A prediction method for asphalt pavement icing based on SVR algorithm was proposed. • A prediction model was established using the SVR method to analyze icing factors. • The BOA was used to automatically adjust the parameters of the prediction model. • The performance of prediction models with different kernel functions was compared. To improve the icing prediction accuracy of asphalt pavement, a prediction method for asphalt pavement icing based on the support vector regression (SVR) algorithm is proposed in this study. A prediction model was established using the SVR method to predict the icing time and thickness of the pavement on bridge at the low solar radiation (rainy), and analyze the effects of the external natural environment (ambient temperature, wind speed and water depth) on the freezing time and icing thickness. The Bayesian optimization algorithm (BOA) was also used to automatically adjust the parameters of the prediction model, which fully considered the coupling correlation of the influencing factors of the asphalt pavement icing. Finally, the prediction performances of the BOA-SVR models with different kernel functions were compared. The results show that the training-set prediction accuracy of the icing time and thickness reaches 99.2% and 92.9%, respectively, and the testing-set prediction accuracy of the icing time and thickness reach 97.7% and 84.4%, respectively. Therefore, the BOA-SVR model has high prediction accuracy. The water depth has the greatest influence on the icing time and thickness of the asphalt pavement, followed by the ambient temperature and wind speed. Overall, the BOA-SVR model can predict the icing time and thickness of the asphalt pavement more precisely compared to existing methods.

Martian subsurface water ice prediction at the Tianwen-1 mission landing site

One of the important sub-tasks of China's ongoing Tianwen-1 exploration mission is to find water ice on Mars. We have predicted the distribution of shallow subsurface water ice at the Tianwen-1 landing site to provide reference information for the mission. A detailed analysis shows that only for an exceptionally favorable combination of slope, albedo, and thermal inertia would perennial near-surface ice be expected at the Tiwanwen-1 landing site and the Zhurong rover is hardly to carry out in-situ searching for ice at steep slopes with dust-mantled surfaces, while seasonal ice is very likely to be found in winter in Tianwen-1 landing area with appropriate terrain conditions. With the interaction of a hygroscopic salt, the transient diurnal ice can exist and experience phase transitions. Notably, under the influence of Ca (ClO4)2, liquid brine may exist in diurnal cycles even on the hottest day of Mars.

Marginal ice zone dynamics: history, definitions and research perspectives.

Despite enormous scientific and technological progress in numerical weather and climate prediction, sea ice still remains unreliably predicted by models, both in short-term forecasting and climate projection applications. The total ice extent in both hemispheres is tied to the location of the ice edge, and consequently to what happens in the portion of the ice cover immediately adjacent to the open ocean that is called the marginal ice zone (MIZ). There is mounting evidence that processes occurring in the MIZ might play an important role in the polar climate of both hemispheres, yet some key physical processes are still missing in models. As sea ice models developed for climate studies are increasingly used for operational forecasting, the missing physics also impede short-term sea ice prediction skills. This paper is a mini-review that provides a historical perspective on how MIZ research has progressed since the 1970s, with a focus on the fundamental importance of the interactions between sea ice and surface gravity waves on sea ice dynamics. Completeness is not achieved, as the body of literature is huge, scattered and rapidly growing, but the intention is to inform future collaborative research efforts to improve our understanding and predictive capabilities of sea ice dynamics in the MIZ. This article is part of the theme issue 'Theory, modelling and observations of marginal ice zone dynamics: multidisciplinary perspectives and outlooks'.

Decadal Sea Ice Prediction in the West Antarctic Seas with Ocean and Sea Ice Initializations

Antarctic sea ice plays an important role in the formation of Antarctic Bottom Water that travels over global oceans and affects global climate. Here we demonstrate that decadal sea ice variability in the west Antarctic seas can be predicted with significant skills using a coupled general circulation model. We present three decadal reforecast experiments where only sea surface temperature, sea surface temperature and sea ice concentration, or sea surface temperature, sea ice concentration and subsurface ocean temperature and salinity are initialized with observations. We find that initializing all three components leads to the highest prediction skills of the sea ice concentration in the Amundsen-Bellingshausen Sea. This experiment captures decadal sea ice increase after the late 2000s, which is linked to anomalous sea ice advection from the Ross Sea and anomalous subsurface ocean cooling by the strengthened Antarctic Circumpolar Current. Skillful prediction of decadal sea ice variability benefits from combined ocean and sea ice initializations.

Mind the gap! A consensus analysis of users and producers on trust in new sea ice information products

In the ice-infested Arctic Ocean environment, the uptake of new sea ice services is an important factor in ensuring safe and efficient marine operations. Producers increasingly turn to co-productionfor user input, similar to the wider field of climate services. This paper asks how the uptake of sea ice information services can be optimized, by gauging the extent to which producers and users alreadyshare an understanding of how trust develops towardnew products. By adopting a consensus analysis approach, we gain insights about how to balance further investments in knowledge co-production versus change implementation. We chose cultural consensus analysis, a method that produces valid estimates even in small sample sizes. Our survey presented thirty-two propositions based on seven dimensions of trust in weather, water, ice and climate services. The survey was completed by fifty-sevenrespondents (n = 29 users, n = 28 producers) and revealed a strong consensus model among the two groups about the necessary improvements needed to increase users’ trust in new services. Our results suggest that forecast producers for the Arctic region, specifically in the field of specialized sea ice predictions and mapping/charting, share a substantial understanding with users about how trust develops toward new products. We discuss the importance of automation, peer endorsement and perceptions of cost-performance ratio for necessary strategic approaches to help experienced forecast users to trust and adapt products to their specific operational context, and reflect on the costs associated with the use of specialized sea ice services in closing the usability gap.

AWI-CM3 coupled climate model: description and evaluation experiments for a prototype post-CMIP6 model

Abstract. We developed a new version of the Alfred Wegener Institute Climate Model (AWI-CM3), which has higher skills in representing the observed climatology and better computational efficiency than its predecessors. Its ocean component FESOM2 (Finite-volumE Sea ice–Ocean Model) has the multi-resolution functionality typical of unstructured-mesh models while still featuring a scalability and efficiency similar to regular-grid models. The atmospheric component OpenIFS (CY43R3) enables the use of the latest developments in the numerical-weather-prediction community in climate sciences. In this paper we describe the coupling of the model components and evaluate the model performance on a variable-resolution (25–125 km) ocean mesh and a 61 km atmosphere grid, which serves as a reference and starting point for other ongoing research activities with AWI-CM3. This includes the exploration of high and variable resolution and the development of a full Earth system model as well as the creation of a new sea ice prediction system. At this early development stage and with the given coarse to medium resolutions, the model already features above-CMIP6-average skills (where CMIP6 denotes Coupled Model Intercomparison Project phase 6) in representing the climatology and competitive model throughput. Finally we identify remaining biases and suggest further improvements to be made to the model.

Prediction of Transmission Line Icing Using Machine Learning Based on GS-XGBoost

In recent years, data have shown that transmission line icing is the main problem affecting the operation of power grids in bad weather; it greatly increases operating costs and affects people’s lives. Therefore, the development of a calculation method to predict the risk of ice on transmission lines is of great importance for the stability of the power grid. In this study, we propose a maximum mutual information coefficient (MIC) and grid search optimization extreme gradient boosting (GS-XGBoost) transmission line ice risk prediction method. First, the MICs between the ice thickness and the precipitation, wind speed, wind direction, relative humidity, slope, aspect, and elevation characteristic factors are calculated to filter out the effective features. Second, a grid search method is used to adjust the hyperparameters of XGBoost. The resulting GS-XGBoost model builds a prediction system based on the best parameters using a training set (70% of the data). Finally, the performance of GS-XGBoost is evaluated using a test set (30% of the data). For multiline, cross-regional icing data, our experimental results show that GS-XGBoost outperforms other machine learning methods in terms of accuracy, precision, recall, and F 1 score.

Study on transmission line icing prediction based on micro-topographic correction

For the requirement of refinement prediction of transmission line icing thickness, this paper presents an icing thickness correction method which combines the prediction using low spatial resolution with the correction of micro-topography features. By analyzing the regional characteristics of micro-topography, the areas can be divided into three categories: plains, high-altitude mountains, and low-altitude mountains. For each type of terrain, the quantitative correction model for the icing thickness is established, which is dependent on altitude, relief amplitude, slope angle, slope aspect, water system, valley line, ridge line, and saddle point. In this paper, according to the numerical prediction on the 3 × 3 km2 grids, the icing thickness of the transmission line on the 300 × 300 m2 grids with micro-topography is obtained using the correction method in Hubei Power Grid. The results show that the correction method is efficient and convenient to improve the accuracy of icing thickness prediction of the transmission line.

Subseasonal-to-Seasonal Arctic Sea Ice Forecast Skill Improvement from Sea Ice Concentration Assimilation

Abstract The current GFDL seasonal prediction system, the Seamless System for Prediction and Earth System Research (SPEAR), has shown skillful prediction of Arctic sea ice extent with atmosphere and ocean constrained by observations. In this study we present improvements in subseasonal and seasonal predictions of Arctic sea ice by directly assimilating sea ice observations. The sea ice initial conditions from a data assimilation (DA) system that assimilates satellite sea ice concentration (SIC) observations are used to produce a set of reforecast experiments (IceDA) starting from the first day of each month from 1992 to 2017. Our evaluation of daily sea ice extent prediction skill concludes that the SPEAR system generally outperforms the anomaly persistence forecast at lead times beyond 1 month. We primarily focus our analysis on daily gridcell-level sea ice fields. SIC DA improves prediction skill of SIC forecasts prominently in the June-, July-, August-, and September-initialized reforecasts. We evaluate two additional user-oriented metrics: the ice-free probability (IFP) and ice-free date (IFD). IFP is the probability of a grid cell experiencing ice-free conditions in a given year, and IFD is the first date on which a grid cell is ice free. A combined analysis of IFP and IFD demonstrates that the SPEAR model can make skillful predictions of local ice melt as early as May, with modest improvements from SIC DA.

Mechanisms of Regional Arctic Sea Ice Predictability in Two Dynamical Seasonal Forecast Systems

Abstract Research over the past decade has demonstrated that dynamical forecast systems can skillfully predict pan-Arctic sea ice extent (SIE) on the seasonal time scale; however, there have been fewer assessments of prediction skill on user-relevant spatial scales. In this work, we evaluate regional Arctic SIE predictions made with the Forecast-Oriented Low Ocean Resolution (FLOR) and Seamless System for Prediction and Earth System Research (SPEAR_MED) dynamical seasonal forecast systems developed at the NOAA/Geophysical Fluid Dynamics Laboratory. Compared to FLOR, we find that the recently developed SPEAR_MED system displays improved skill in predicting regional detrended SIE anomalies, partially owing to improvements in sea ice concentration (SIC) and thickness (SIT) initial conditions. In both systems, winter SIE is skillfully predicted up to 11 months in advance, whereas summer minimum SIE predictions are limited by the Arctic spring predictability barrier, with typical skill horizons of roughly 4 months. We construct a parsimonious set of simple statistical prediction models to investigate the mechanisms of sea ice predictability in these systems. Three distinct predictability regimes are identified: a summer regime dominated by SIE and SIT anomaly persistence; a winter regime dominated by SIE and upper-ocean heat content (uOHC) anomaly persistence; and a combined regime in the Chukchi Sea, characterized by a trade-off between uOHC-based and SIT-based predictability that occurs as the sea ice edge position evolves seasonally. The combination of regional SIE, SIT, and uOHC predictors is able to reproduce the SIE skill of the dynamical models in nearly all regions, suggesting that these statistical predictors provide a stringent skill benchmark for assessing seasonal sea ice prediction systems.

Prediction of Pan-Arctic Sea Ice Using Attention-Based LSTM Neural Networks

Within the rapidly changing Arctic region, accurate sea ice forecasts are of crucial importance for navigation activities, such as the planning of shipping routes. Numerical climate models have been widely used to generate Arctic sea ice forecasts at different time scales, but they are highly dependent on the initial conditions and are computationally expensive. Recently, with the increasing availability of geoscience data and the advances in deep learning algorithms, the use of artificial intelligence (AI)-based sea ice prediction methods has gained significant attention. In this study, we propose a supervised deep learning approach, namely attention-based long short-term memory networks (LSTMs), to forecast pan-Arctic sea ice at monthly time scales. Our method makes use of historical sea ice concentration (SIC) observations during 1979–2020, from passive microwave brightness temperatures. Based on the persistence of SIC anomalies, which is known as one of the dominant sources of sea ice predictability, our approach exploits the temporal relationships of sea ice conditions across different time windows of the training period. We demonstrate that the attention-based LSTM is able to learn the variations of the Arctic sea ice and can skillfully forecast pan-Arctic SIC on monthly time scale. By designing the loss function and utilizing the attention mechanism, our approach generally improves the accuracy of sea ice forecasts compared to traditional LSTM networks. Moreover, it outperforms forecasts with the climatology and persistence based empirical models, as well as two dynamical models from the Copernicus Climate Change Service (C3S) datastore. This approach shows great promise in enhancing forecasts of Arctic sea ice using AI methods.

Geospatial Simulation System of Mountain Area Black Ice Accidents

As the development of mountain areas has recently increased in Korea, existing roads are being renovated, and new highways are being constructed, which increases driving speeds in mountainous areas. However, the mountainous region in northeastern Korea is more likely to form black ice due to higher humidity, frequent fog, and hillshade, depending on the terrain, which can cause serious traffic pileups. In this study, therefore, we present a method to build a more effective black ice prediction and warning system by linking spatial information to the existing road management system that estimates the road surface temperature based on real-time weather information. The spatial information enabled a prediction to be made of the risk level of black ice formation for each time zone by simulating changes in the shadow area based on precise 3D terrain information. Moreover, this information also presented slope and curvature information of the road to estimate the risk zone. The spatial information was integrated with weather data to predict road surface temperature. The proposed system was tested in two mountainous regions with weather data accumulated from 2017 to 2018. As a result, the proposed system anticipated 71% of traffic accidents caused by black ice during the testing period. The results show that the system can contribute significantly to preventing black-ice-related traffic accidents by providing reasonable predictions.

ИСПОЛЬЗОВАНИЕ СИСТЕМЫ COSMO-RU ДЛЯ ПРОГНОЗА ОБЛЕДЕНЕНИЯ ВОЗДУШНЫХ СУДОВ НАД ТЕРРИТОРИЕЙ РОССИЙСКОЙ ФЕДЕРАЦИИ

A method is proposed for in-flight icing prediction up to 12 hours based on the output of the COSMO-Ru configuration of the COSMO model with a grid step of 6.6 km for the Russian Federation area. The calculation of the simplified forecast icing potential (SFIP) using model microphysics data versus AIREP Speci information allowed determining a threshold value of SFIP ≥ 0.2, the exceeding of which is an icing criterion. A good agreement was found between the icing zones obtained by SFIP and the areas identified according to the condition -16°C ≤ T ≤ 0°C, RH ≥ 85%. To distinguish severe icing, it is proposed to use the values of model liquid water content exceeding 0.4 g/kg.

Application of the COSMO-Ru System for Aircraft Icing Prediction over the Russian Federation Area

Application of the COSMO-Ru System for Aircraft Icing Prediction over the Russian Federation Area

Quantifying the Contribution of Internal Atmospheric Drivers to Near-Term Projection Uncertainty in September Arctic Sea Ice

Abstract Arctic sea ice has undergone rapid loss in all months of the year in recent decades, especially in September. The September sea ice extent (SSIE) in the multimodel ensemble mean of climate models shows a large divergence from observations since the 2000s, which indicates the potential influence of internal variability on SSIE decadal variations. Reasons previously identified for the accelerated decrease in SSIE are largely related to the tendency toward a barotropic geopotential height rise in summer over the Arctic. We used a 40-member ensemble of simulation by the Community Earth System Model version 1 (CESM1) and a 100-member ensemble simulation by the Max Planck Institute Earth System Model (MPI-ESM) to reveal that the internal variability of the local atmosphere circulation change can contribute 12%–17% to the uncertainties in the projected SSIE changes during 2016–45 in both CESM-LE and MPI-ESM. The tropical Pacific Ocean may act as a remote driver for the sea ice melting but the coupling between them is more intense on decadal time scales than that on year-to-year scales. Our quantitative estimation of the contribution of the internal atmospheric circulation to SSIE during the next three decades may be underestimated due to models’ inability to capture the observed Rossby wave train originating from the tropical Pacific Ocean propagating into the Arctic. Further efforts toward investigating causes of the model limitations and quantifying the contribution of local and remote component to Arctic sea ice on different time scales may help to improve the future sea ice prediction.

Research on Dynamic Assessment System of Composite Fault Risk of Transmission Line Based on Blockchain Energy

In order to study the dynamic assessment system of composite fault risk of transmission line based on blockchain energy and in order to study the transmission line compound fault risk dynamic assessment system based on blockchain, firstly, according to the coupling relationship between power grid and natural disasters, the information resources such as data collected by power grid intelligent devices and natural meteorology are excavated, and the overall architecture of power grid disaster early warning and decision-making system supported by blockchain is built. Then, from the perspective of risk, combined with analytic hierarchy process, an index system for reasonable evaluation of distribution network fault benchmark risk is established. Quantitative assessment and risk classification shall be carried out for the failure probability, failure impact consequence, and comprehensive failure risk, so as to facilitate the adoption of risk response measures. Finally, taking several 220 kV lines in the northwest and central part of a city as examples, the icing prediction analysis verifies the feasibility and effectiveness of the proposed power grid disaster early warning decision system based on blockchain to predict the icing thickness. The experimental results show that taking the icing disaster as an example, the MPC method is used to modify the icing thickness prediction model, improve the accuracy of the icing prediction model, and verify the feasibility and effectiveness of the prediction and early warning system based on blockchain.