Short-term Forecasting System Research Articles

Forecasting the Mediterranean Sea marine heatwave of summer 2022

Abstract. Early warning of marine heatwaves requires short-term forecasts to provide precise information on timings, local-scale coverage, and intensities of coming events. Here, we describe our successful efforts to track the onset, peak, and decay of the Mediterranean Sea marine heatwave of summer 2022 with the Copernicus MedFS short-term (10 d) forecast system. First, we show that the 2022 event eclipses the economically and ecologically damaging event of 2003 in terms of marine heatwave (MHW) activity (a measure of intensity and duration). Forecasts of MHW area and activity provide a means of basin-wide validation, highlighting the capability of MedFS to capture regional behaviour. On local scales, we found that the MHW occurrence in the Ligurian Sea and Gulf of Taranto, two regions of economic and ecological importance, was also reliably forecast. Encouragingly, we note that the forecast has demonstrated skill in capturing not just the season-long MHW cycle but also breaks in MHW persistence and abrupt changes in local activity. Subseasonal forecasts do not yet demonstrate the capacity to predict MHW response to short-lived weather patterns, but this study confirms that short-term forecasts, at least in the Mediterranean Sea, can fill this gap.

Evaluation of the CMCC global eddying ocean model for the Ocean Model Intercomparison Project (OMIP2)

Abstract. This paper describes the global eddying ocean–sea ice simulation produced at the Euro-Mediterranean Center on Climate Change (CMCC) obtained following the experimental design of the Ocean Model Intercomparison Project phase 2 (OMIP2). The eddy-rich model (GLOB16) is based on the NEMOv3.6 framework, with a global horizontal resolution of 1/16∘ and 98 vertical levels and was originally designed for an operational short-term ocean forecasting system. Here, it is driven by one multi-decadal cycle of the prescribed JRA55-do atmospheric reanalysis and runoff dataset in order to perform a long-term benchmarking experiment. To assess the accuracy of simulated 3D ocean fields and highlight the relative benefits of resolving mesoscale processes, the GLOB16 performances are evaluated via a selection of key climate metrics against observational datasets and two other NEMO configurations at lower resolutions: an eddy-permitting resolution (ORCA025) and a non-eddying resolution (ORCA1) designed to form the ocean–sea ice component of the fully coupled CMCC climate model. The well-known biases in the low-resolution simulations are significantly improved in the high-resolution model. The evolution and spatial pattern of large-scale features (such as sea surface temperature biases and winter mixed-layer structure) in GLOB16 are generally better reproduced, and the large-scale circulation is remarkably improved compared to the low-resolution oceans. We find that eddying resolution is an advantage in resolving the structure of western boundary currents, the overturning cells, and flow through key passages. GLOB16 might be an appropriate tool for ocean climate modeling efforts, even though the benefit of eddying resolution does not provide unambiguous advances for all ocean variables in all regions.

AUnified Graph Formulation for Spatio-Temporal Wind Forecasting

With the rapid adoption of wind energy globally, there is a need for accurate short-term forecasting systems to improve the reliability and integration of such energy resources on a large scale. While most spatio-temporal forecasting systems comprise distinct components to learn spatial and temporal dependencies separately, this paper argues for an approach to learning spatio-temporal information jointly. Many time series forecasting systems also require aligned input information and do not naturally facilitate irregular data. Research is therefore required to investigate methodologies for forecasting in the presence of missing or corrupt measurements. To help combat some of these challenges, this paper studied a unified graph formulation. With the unified formulation, a graph neural network (GNN) was used to extract spatial and temporal dependencies simultaneously, in a single update, while also naturally facilitating missing data. To evaluate the proposed unified approach, the study considered hour-ahead wind speed forecasting in the North Sea under different amounts of missing data. The framework was compared against traditional spatio-temporal architectures that used GNNs together with temporal long short-term memory (LSTM) and Transformer or Autoformer networks, along with the imputation of missing values. The proposed framework outperformed the traditional architectures, with absolute errors of around 0.73–0.90 m per second, when subject to 0–80% of missing input data. The unified graph approach was also better at predicting large changes in wind speed, with an additional 10-percentage-point improvement over the second-best model. Overall, this paper investigated a novel methodology for spatio-temporal wind speed forecasting and showed how the proposed unified graph formulation achieved competitive results compared to more traditional GNN-based architectures.

Modal analysis of lead-bismuth eutectic flow in a single wire-wrapped rod channel

The thermal-hydraulics as well as flow-induced vibration of wire-wrapped rod bundles calls for accurate and efficient liquid metal flow simulation and prediction, yet it remains a challenge due to the complex geometries and high Reynolds number flow in wire-wrapped rod channel. Previous efforts towards this goal exclusively adopts full-order modeling (FOM), which is prohibitively computation-intensive. This work reports the first reference on modal analysis of the turbulent lead-bismuth eutectic (LBE) flow in a single wire-wrapped rod channel by a data-driven dynamic mode decomposition (DMD) method. The spatio-temporal data of fluid velocities and pressure were modeled and collected by using the Reynolds-averaged Navier–Stokes equations (RANS) approach with stress-omega RSM turbulence model. Modal analysis was performed over the snapshot matrices and DMD modes and corresponding eigenvalues were calculated. The DMD modes and eigenvalues were then used to construct a reduced-order modeling (ROM) system for fast fluid field reconstruction and short-term forecasting. The correctness and feasibility of the method was demonstrated and compared with ground-truth CFD simulation results. Our method offers a fast and reliable approach for ROM of turbulent flow in wire-wrapped rod fuel assemblies. We publicize all the data and codes via https://github.com/XJTU-Zhou-group/Wire-wrapped-fuel-pin-CFD-DMD.

ВОЗМОЖНОСТИ ИСПОЛЬЗОВАНИЯ СИСТЕМЫ COSMO-RU ПРИ КРАТКОСРОЧНОМ ПРОГНОЗИРОВАНИИ СТОКА РЕК РОССИИ

An assessment of the possibilities of using the COSMO-Ru system for short-term forecasting of runoff of Russian rivers is carried out. An analysis of the forecast quality of the daily precipitation layer and average daily temperature of the surface air layer using this system is performed for 2081 weather stations. The quality characteristics of these forecasts are mapped. A forecast correction method was applied, which made it possible to increase the accuracy of forecasts on average by 20% for the precipitation layer and by 15% for air temperature. The quality of forecasting meteorological parameters averaged over the territory of 590 river basins turned out to be satisfactory for the precipitation layer and good for air temperature. The results give reason for a positive assessment of the possibilities of using the COSMO-Ru system for short-term forecasting of runoff of Russian rivers.

Possibilities of Using the COSMO-Ru System for Short-term Forecasting of Runoff of Russian Rivers

Possibilities of Using the COSMO-Ru System for Short-term Forecasting of Runoff of Russian Rivers

The Texas A&M – University of Haifa Eastern Mediterranean Observatory: Monitoring the Eastern Mediterranean Sea

The need for sustained long-term measurements of the ocean has increased due to climate variability and societal desire to mitigate the impact of natural disasters on coastal communities. Knowledge about key ocean variables is essential for monitoring and predicting long-term ocean variability, and real-time capability is needed to inform short-term forecasting systems to assist coastal managers and response teams. Ocean observatories collect oceanographic data for scientific and societal reasons, and open access to these data enables their use by the general public, decision-makers, and stakeholders.

Generating a long-term (2003−2020) hourly 0.25° global PM2.5 dataset via spatiotemporal downscaling of CAMS with deep learning (DeepCAMS)

Generating a long-term high-spatiotemporal resolution global PM2.5 dataset is of great significance for environmental management to mitigate the air pollution concerns worldwide. However, the current long-term (2003–2020) global reanalysis dataset Copernicus Atmosphere Monitoring Service (CAMS) reanalysis has drawbacks in fine-scale research due to its coarse spatiotemporal resolution (0.75°, 3-h). Hence, this paper developed a deep learning-based framework (DeepCAMS) to downscale CAMS PM2.5 product on the spatiotemporal dimension for resolution enhancement. The nonlinear statistical downscaling from low-resolution (LR) to high-resolution (HR) data can be learned from the high quality (0.25°, hourly) but short-term (2018–2020) Goddard Earth Observing System composition forecast (GEOS-CF) system PM2.5 product. Compared to the conventional spatiotemporal interpolation methods, simulation validations on GEOS-CF demonstrate that DeepCAMS is capable of producing accurate temporal variations with an improvement of Root-Mean-Squared Error (RMSE) of 0.84 (4.46 to 5.30) ug/m3 and spatial details with an improvement of Mean Absolute Error (MAE) of 0.16 (0.34 to 0.50) ug/m3. The real validations on CAMS reflect convincing spatial consistency and temporal continuity at both regional and global scales. Furthermore, the proposed dataset is validated with OpenAQ air quality data from 2017 to 2019, and the in-situ validations illustrate that the DeepCAMS maintains the consistent precision (R: 0.597) as the original CAMS (R: 0.593) while tripling the spatiotemporal resolution. The proposed dataset will be available at https://doi.org/10.5281/zenodo.6381600.

Application of Digital Twin Technology in Auxiliary Decision-making System for Grid-connected Dispatching of New Energy

Digital twin is the digital expression of physical entities. With the further development of big data and artificial intelligence technology, many fields are actively promoting the digital transformation of industries. As the key area of Chinese infrastructure to promote clean coal production and efficient green transformation, using the digital twin technology to promote the overall perception of new energy power stations, intelligent decision-making and dynamic prediction has become a new research hotspot. Through the intelligent acquisition in the perception layer, intelligent integration in the data layer and intelligent decision making in the service layer, we can set up a whole set of super short-term forecasting system and dispatching decision support system for new energy power stations, which can improve the operational efficiency. From hindsight to foresight, It’s an innovative imporve to promote the construction of smart power grid and digital power grid by effectively solving the abandon of wind and irradiance, reducing the cost of power generation and ensuring the stable operation of the whole power grid.

Research of a novel short-term wind forecasting system based on multi-objective Aquila optimizer for point and interval forecast

Facing the increasing depletion of traditional energy resources and the worsening environmental issues, wind energy sources have been widely considered. As an essential renewable energy resource, wind energy features abundant deposits, extensive distribution, non-pollution, etc. In recent years, wind power generation occupies a non-negligible position in the electric power industry. Stable and reliable power system operation demands accurate wind speed prediction (WSP), but the inherent randomness of wind speed sequences complicates their fluctuations and causes them to be uncontrollable. In this paper, an innovative WSP system is proposed, which combines data pre-processing technique, benchmark model selection, an advanced optimizer for point forecast and interval forecast. Furthermore, this paper theoretically demonstrates that the weights allocated by this optimizer are Pareto optimal solutions. Six interval data from two sites in China are utilized to validate the forecasting performance of our developed model. The experimental results indicate that the developed model can achieve superior accuracy compared to the tested models in all cases for point forecast, and also obtains the forecasting interval with high coverage and low width error, which is an extremely crucial instruction to guarantee the security and stability of the power system.

Development of a Low-Cost Data Acquisition System for Very Short-Term Photovoltaic Power Forecasting

The rising adoption of renewable energy sources means we must turn our eyes to limitations in traditional energy systems. Intermittency, if left unaddressed, may lead to several power-quality and energy-efficiency issues. The objective of this work is to develop a working tool to support photovoltaic energy forecast models for real-time operation applications. The current paradigm of intra-hour solar-power forecasting is to use image-based approaches to predict the state of cloud composition for short time horizons. Since the objective of intra-minute forecasting is to address high-frequency intermittency, data must provide information on and surrounding these events. For that purpose, acquisition by exception was chosen as the guiding principle. The system performs power measurements at 1 Hz frequency, and whenever it detects variations over a certain threshold, it saves the data 10 s before and 4 s after the detection point. A multilayer perceptron neural network was used to determine its relevance to the forecasting problem. With a thorough selection of attributes and network structures, the results show very low error with R2 greater than 0.93 for both input variables tested with a time horizon of 60 s. In conclusion, the data provided by the acquisition system yielded relevant information for forecasts up to 60 s ahead.

Application of a Short-Term Hydrological Forecast System

Bu çalışmanın amacı kar erimesinin etkili olduğu dağlık memba havzaları için kısa dönemli akım tahmin sisteminin geliştirilmesidir. Çalışma alanları olarak seçilen dağlık Fırat ve Seyhan Havzaları, yüksek su potansiyeli, bu potansiyeli besleyen kar erimeleri, mansapta büyük ve önemli su yapılarının bulunması ile ön plana çıkmaktadırlar. Yağış-akış ilişkisinin simülasyonu için dağlık bölge uygulamalarında literatürde yaygın olarak kullanılan HBV modeli seçilmiştir. Sayısal Hava Tahmin (SHT) verisi olarak Mesoscale Model 5 (MM5) ve Weather Research and Forecast (WRF) model sonuçları kullanılmıştır. Hidrolojik model parametrelerinin gözlenen yağış, sıcaklık ve akım verileriyle kalibrasyon/doğrulama işleminin yapılmasının ardından, analizleri yapılan SHT verilerinin girdi olarak kullanılması ile 1 ve 2 günlük akım tahminleri elde edilmiştir. İleriye dönük akım tahminleri Delft-FEWS platformunda kapalı döngü şeklinde çalıştırılarak, geçmiş dönem tahmin simülasyonları gerçekleştirilmiş ve akım gözlemleri ile kıyaslanarak performans değerlendirmesi yapılmıştır. Elde edilen sonuçlar havzaların akış aşağısında bulunan su yapılarının daha verimli işletilmesine ve böylece ülke ekonomisine katkı sağlayabilecektir.

Short-Term Forecasting of Large-Scale Clouds Impact on Downwelling Surface Solar Irradiation

This study focuses on the use of cloud motion vectors (CMV) and fast radiative transfer models (FRTM) in the prospect of forecasting downwelling surface solar irradiation (DSSI). Using near-real-time cloud optical thickness (COT) data derived from multispectral images from the spinning enhanced visible and infrared imager (SEVIRI) onboard the Meteosat second generation (MSG) satellite, we introduce a novel short-term forecasting system (3 h ahead) that is capable of calculating solar energy in large-scale (1.5 million-pixel area covering Europe and North Africa) and in high spatial (5 km over nadir) and temporal resolution (15 min intervals). For the operational implementation of such a big data computing architecture (20 million simulations in less than a minute), we exploit a synergy of high-performance computing and deterministic image processing technologies (dense optical flow estimation). The impact of clouds forecasting uncertainty on DSSI was quantified in terms of cloud modification factor (CMF), for all-sky and clear sky conditions, for more generalized results. The forecast accuracy was evaluated against the real COT and CMF images under different cloud movement patterns, and the correlation was found to range from 0.9 to 0.5 for 15 min and 3 h ahead, respectively. The CMV forecast variability revealed an overall DSSI uncertainty in the range 18–34% under consecutive alternations of cloud presence, highlighting the ability of the proposed system to follow the cloud movement in opposition to the baseline persistent forecasting, which considers the effects of topocentric sun path on DSSI but keeps the clouds in “fixed” positions, and which presented an overall uncertainty of 30–43%. The proposed system aims to support the distributed solar plant energy production management, as well as electricity handling entities and smart grid operations.

The Impact of Data Filtration on the Accuracy of Multiple Time-Domain Forecasting for Photovoltaic Power Plants Generation

The paper reports the forecasting model for multiple time-domain photovoltaic power plants, developed in response to the necessity of bad weather days’ accurate and robust power generation forecasting. We provide a brief description of the piloted short-term forecasting system and place under close scrutiny the main sources of photovoltaic power plants’ generation forecasting errors. The effectiveness of the empirical approach versus unsupervised learning was investigated in application to source data filtration in order to improve the power generation forecasting accuracy for unstable weather conditions. The k-nearest neighbors’ methodology was justified to be optimal for initial data filtration, based on the clusterization results, associated with peculiar weather and seasonal conditions. The photovoltaic power plants’ forecasting accuracy improvement was further investigated for a one hour-ahead time-domain. It was proved that operational forecasting could be implemented based on the results of short-term day-ahead forecast mismatches predictions, which form the basis for multiple time-domain integrated forecasting tools. After a comparison of multiple time series forecasting approaches, operational forecasting was realized based on the second-order autoregression function and applied to short-term forecasting errors with the resulting accuracy of 87%. In the concluding part of the article the authors from the points of view of computational efficiency and scalability proposed the hardware system composition.

Industry Experience of Developing Day-Ahead Photovoltaic Plant Forecasting System Based on Machine Learning

This article highlights the industry experience of the development and practical implementation of a short-term photovoltaic forecasting system based on machine learning methods for a real industry-scale photovoltaic power plant implemented in a Russian power system using remote data acquisition. One of the goals of the study is to improve photovoltaic power plants generation forecasting accuracy based on open-source meteorological data, which is provided in regular weather forecasts. In order to improve the robustness of the system in terms of the forecasting accuracy, we apply newly derived feature introduction, a factor obtained as a result of feature engineering procedure, characterizing the relationship between photovoltaic power plant energy production and solar irradiation on a horizontal surface, thus taking into account the impacts of atmospheric and electrical nature. The article scrutinizes the application of different machine learning algorithms, including Random Forest regressor, Gradient Boosting Regressor, Linear Regression and Decision Trees regression, to the remotely obtained data. As a result of the application of the aforementioned approaches together with hyperparameters, tuning and pipelining of the algorithms, the optimal structure, parameters and the application sphere of different regressors were identified for various testing samples. The mathematical model developed within the framework of the study gave us the opportunity to provide robust photovoltaic energy forecasting results with mean accuracy over 92% for mostly-sunny sample days and over 83% for mostly cloudy days with different types of precipitation.

Automated short-term forecast system based on open-source hydrological models for the Tikhvinka river (Leningrad region of Russia)

In recent decades there has been a trend towards an increase in the number of dangerous hydrological events, especially floods. In order to protect citizens and solve economic problems, it is important to develop and actively introduce into operational practice methods of hydrological forecasting, as well as to build more modern and convenient interfaces of interaction between hydrometeorological services, municipal authorities and citizens. This work discusses a compact automated short-term hydrological forecasting system that uses open-source conceptual models HBV, SimHYD and GR4J as its core. The system is connected to data streams on the observed temperatures and precipitation in the watershed basin, as well as the predicted values of these parameters (in a current implementation, the WRF model with a forecast for 84 hours is used). Also, for operational calibration in daily mode, the system can assimilate (if available) data on observed water levels. Testing of the system is carried out on the example of Tikhvin city (the Tikhvinka river), which in recent years has been characterized by frequent flooding.

Deep learning in exchange markets

We present the implementation of a short-term forecasting system of price movements in exchange markets using market depth data and a systematic procedure to enable a fully automated trading system. Three types of Deep Learning (DL) Neural Network (NN) methodologies are trained and tested: Deep NN Classifier (DNNC), Long Short-Term Memory (LSTM) and Convolutional NN (CNN). Although the LSTM is more suitable for multivariate time series analysis from a theoretical point of view, test results indicate that the CNN has on average the best predictive power in the case study under analysis, which is the UK to Win Horse Racing market during pre-live stage in the world’s most relevant betting exchange. Implications from the generalized use of automated trading systems in betting exchange markets are discussed.

Forecasting severe convective storms with WRF-based RTFDDA radar data assimilation in Guangdong, China

Radar data assimilation is an important method for short-term convection forecasting or nowcasting. To improve the short-term (mainly 0–3 h) precipitation forecasts for severe convective storms, an analysis nudging (Newtonian relaxation) based hydrometeor and latent heat nudging (HLHN) technique was developed to effectively assimilate radar reflectivity data in a Weather Research and Forecasting (WRF)-based real time four-dimensional data assimilation and short-term forecasting system (RTFDDA). The purpose of this study is to investigate the performance of the RTFDDA system with radar data assimilation (RTFDDA-RDA) with rapid-cycling forecasting applications for Shenzhen, a subtropical coastal metropolis in southern China. The RTFDDA-RDA system was run to produce hindcasts for ten severe convective storm events occurred in Guangdong region during the 2017 rainy season. Results show that, through nudging cloud hydrometeors retrieved from radar reflectivity and the associated latent heat release, RTFDDA-RDA is able to produce the meso- and convective-scale features of the convective storms in a good accuracy and improve the short-term precipitation forecasting of the convective storms. Subjective and statistical evaluation results demonstrate that RTFDDA-RDA presents a reasonable capability for forecasting convective systems with improving the initial conditions and resulting in significant improvements of precipitation forecasting skills, especially for the 0–3-h nowcasting range. The sensitivity experiments on different latent heating schemes show that, the convective-stratiform separated heating scheme has the best performance of forecasts. Finally, intercomparison of different radar data assimilation approaches will be conducted in future.

Solar power generation short-term forecasting model’s implementation experience

Recently there is significant increase in the installed capacity of solar power plants in Russia. Thereby there are issues of solar power plants owners information support for participation in wholesale electricity market. The paper describes the experience of short term forecasting system practical implementation. The system is proposed for forecasting the solar power plant generation “a day ahead” as a part of the software for automatic meter reading systems “Energosfera”. The short-term forecasting program modules structure, key parameters and characteristics used during the forecasting process description is presented.

A finite element multigrid-framework to solve the sea ice momentum equation

We present a multigrid framework for the solution of the momentum equation arising in Hibler's viscous-plastic (VP) sea ice model. This model is used in global climate models, for seasonal prediction and high-resolution short-term forecasting systems. The development of fast and robust solvers for the strongly nonlinear momentum equation is still a big issue. There are mainly three approaches to solve the momentum equation: a fixed-point iteration (Picard solver), an inexact Newton method and a subcycling procedure based on an elastic-viscous-plastic (EVP) model approximation. Simple fixed-point iterations call for a vast number of cycles. The Jacobian arising in Newton linearizations is ill-conditioned and unstructured. No efficient linear solvers are available up to date. One possibility to solve the linear systems arising within a Newton method is the preconditioned GMRES iteration that however still requires many steps. The commonly used line SOR preconditioner is computationally expensive. Especially on fine meshes, literature recommends to revise common solution strategies.We introduce a geometric multigrid method as a preconditioner to the GMRES iteration for accelerating the solution of the linear problems. We show that the convergence rate of the multigrid method is robust with respect to mesh refinement. This makes it an appealing method for high resolution simulations. We validate the robustness of the linear solver and compare the multigrid with ILU preconditioning. In particular on fine meshes (∼16 km–2 km), multigrid preconditioning can substantially reduce the computational effort and decreases iteration counts by 80%.