Benefits of smoothing backgrounds and radar reflectivity observations for multiscale data assimilation with an ensemble Kalman filter at convective scales: A proof of concept study

In the ensemble Kalman filter (EnKF), the covariance localization radius is usually small when assimilating radar observations because of high density of the radar observations. This makes the region away from precipitation difficult to correct if no other observations are available, as there is no reason to correct the background. To correct errors away from the innovating radar observations, a multiscale localization (MLoc) method adapted to dense observations like those from radar is proposed. In this method, different scales are corrected successively by using the same reflectivity observations, but with different degree of smoothing and localization radius at each step. In the context of observing system simulation experiments, single and multiple assimilation experiments are conducted with the MLoc method. Results show that the MLoc assimilation updates areas that are away from the innovative observations and improves on average the analysis and forecast quality in single cycle and cycling assimilation experiments. The forecast gains are maintained until the end of the forecast period, illustrating the benefits of correcting different scales.

Quality of the Wind Wave Forecast in the Black Sea Including Storm Wave Analysis

This paper presents the results of wind wave forecasts for the Black Sea. Three different versions utilized were utilized: the WAVEWATCH III model with GFS 0.25 forcing on a regular grid, the WAVEWATCH III model with COSMO-RU07 forcing on a regular grid, and the SWAN model with COSMO-RU07 forcing on an unstructured grid. AltiKa satellite altimeter data were used to assess the quality of wind and wave forecasts for the period from 1 April to 31 December 2017. Wave height and wind speed forecast data were obtained with a lead time of up to 72 h. The presented models provide an adequate forecast in terms of modern wave modeling (a correlation coefficient of 0.8–0.9 and an RMSE of 0.25–0.3 m) when all statistics were analyzed. A clear improvement in the wave forecast quality with the high-resolution wind forecast COSMO-RU07 was not registered. The bias error did not exceed 0.5 m in an SWH range from 0 to 3 m. However, the bias sharply increased to −2 or −3 m for an SWH range of 3–4 m. Wave forecast quality assessments were conducted for several storm cases.

AI-based techniques for multi-step streamflow forecasts: application for multi-objective reservoir operation optimization and performance assessment

Streamflow forecasts are traditionally effective in mitigating water scarcity and flood defense. This study developed an artificial intelligence (AI)-based management methodology that integrated multi-step streamflow forecasts and multi-objective reservoir operation optimization for water resource allocation. Following the methodology, we aimed to assess forecast quality and forecast-informed reservoir operation performance together due to the influence of inflow forecast uncertainty. Varying combinations of climate and hydrological variables were input into three AI-based models, namely a long short-term memory (LSTM), a gated recurrent unit (GRU), and a least-squares support vector machine (LSSVM), to forecast short-term streamflow. Based on three deterministic forecasts, the stochastic inflow scenarios were further developed using Bayesian model averaging (BMA) for quantifying uncertainty. The forecasting scheme was further coupled with a multi-reservoir optimization model, and the multi-objective programming was solved using the parameterized multi-objective robust decision-making (MORDM) approach. The AI-based management framework was applied and demonstrated over a multi-reservoir system (25 reservoirs) in the Zhoushan Islands, China. Three main conclusions were drawn from this study: (1) GRU and LSTM performed equally well on streamflow forecasts, and GRU might be the preferred method over LSTM, given that it had simpler structures and less modeling time; (2) higher forecast performance could lead to improved reservoir operation, while uncertain forecasts were more valuable than deterministic forecasts, regarding two performance metrics, i.e., water supply reliability and operating costs; (3) the relationship between the forecast horizon and reservoir operation was complex and depended on the operating configurations (forecast quality and uncertainty) and performance measures. This study reinforces the potential of an AI-based stochastic streamflow forecasting scheme to seek robust strategies under uncertainty.

Artificial Neural Network Based shafts surface pressures Analysis

This paper exhibits a development of an Artificial Neural Network (ANN) as an instrument for investigation of various parameters of a framework. ANN comprises of various layers of straightforward handling components called as neurons. The neuron performs two capacities, to be specific, assortment of sources of info and age of a yield. Utilization of ANN gives diagram of the hypothesis, learning rules, and uses of the most significant neural system models, definitions and style of Computation. The scientific model of system illuminates the idea of sources of info, loads, adding capacity, actuation work and yields. At that point ANN chooses the sort of learning for modification of loads with change in parameters. At long last the examination of a framework is finished by ANN execution and ANN preparing and forecast quality.

Time series forecasting using amplitude-frequency analysis of STL components

The problem of economic time series analysis and forecasting using amplitude-frequency analysis of STL decomposition is considered. An amplitude-phase operator was chosen as an apparatus for extraction the series harmonic components, the advantages of which (compared to the Fourier transform) are: calculations speed, result accuracy, simplicity and interpretability of software implementation. The forecast quality was carried out using the MAPE metric. Significantly higher prediction quality was achieved compared to Facebook Prophet forecasting package.

Devising a method for constructing the optimal model of time series forecasting based on the principles of competition

This paper reports the analysis of a forecasting problem based on time series. It is noted that the forecasting stage itself is preceded by the stages of selection of forecasting methods, determining the criterion for the forecast quality, and setting the optimal prehistory step. As one of the criteria for a forecast quality, its volatility has been considered. Improving the volatility of the forecast could ensure a decrease in the absolute value of the deviation of forecast values from actual data. Such a criterion should be used in forecasting in medicine and other socially important sectors. To implement the principle of competition between forecasting methods, it is proposed to categorize them based on the values of deviations in the forecast results from the exact values of the elements of the time series. The concept of dominance among forecasting methods has been introduced; rules for the selection of dominant and accurate enough predictive models have been defined. Applying the devised rules could make it possible, at the preceding stages of the analysis of the time series, to reject in advance the models that would surely fail from the list of predictive models available to use. In accordance with the devised method, after applying those rules, a system of functions is built. The functions differ in the sets of predictive models whose forecasting results are taken into consideration. Variables in the functions are the weight coefficients with which predictive models are included. Optimal values for the variables, as well as the optimal model, are selected as a result of minimizing the functions built. The devised method was experimentally verified. It has been shown that the constructed method made it possible to reduce the forecast error from 0.477 and 0.427 for basic models to 0.395 and to improve the volatility of the forecast from 1969.489 and 1974.002 to 1607.065

Data Assimilation of Volcanic Aerosols using FALL3D+PDAF

Modelling atmospheric dispersal of volcanic ash and aerosols is becoming increasingly valuable for assessing the potential impacts of explosive volcanic eruptions on infrastructures, air quality, and aviation. Management of volcanic risk and reduction of aviation impacts can strongly benefit from quantitative forecasting of volcanic ash. However, an accurate prediction of volcanic aerosol concentrations using numerical modelling relies on proper estimations of multiple model parameters which are prone to errors. Uncertainties in key parameters such as eruption column height, physical properties of particles or meteorological fields, represent a major source of error affecting the forecast quality. The availability of near-real-time geostationary satellite observations with high spatial and temporal resolutions provides the opportunity to improve forecasts in an operational context by incorporating observations into numerical models. Specifically, ensemble-based filters aim at converting a prior ensemble of system states into an analysis ensemble by assimilating a set of noisy observations. Previous studies dealing with volcanic ash transport have demonstrated that a significant improvement of forecast skill can be achieved by this approach. In this work, we present a new implementation of an ensemble-based Data Assimilation (DA) method coupling the FALL3D dispersal model and the Parallel Data Assimilation Framework (PDAF). The FALL3D+PDAF system runs in parallel, supports online-coupled DA and can be efficiently integrated into operational workflows by exploiting high-performance computing (HPC) resources. Two numerical experiments are considered: (i) a twin experiment using an incomplete dataset of synthetic observations of volcanic ash and, (ii) an experiment based on the 2019 Raikoke eruption using real observations of SO2 mass loading. An ensemble-based Kalman filtering technique based on the Local Ensemble Transform Kalman Filter (LETKF) is used to assimilate satellite-retrieved data of column mass loading. We show that this procedure may lead to nonphysical solutions and, consequently, conclude that LETKF is not the best approach for the assimilation of volcanic aerosols. However, we find that a truncated state constructed from the LETKF solution approaches the real solution after a few assimilation cycles, yielding a dramatic improvement of forecast quality when compared to simulations without assimilation.

The smoother the better? A comparison of six post-processing methods to improve short-term offshore wind power forecasts in the Baltic Sea

With a rapidly increasing capacity of electricity generation from wind power, the demand for accurate power production forecasts is growing. To date, most wind power installations have been onshore and thus most studies on production forecasts have focused on onshore conditions. However, as offshore wind power is becoming increasingly popular it is also important to assess forecast quality in offshore locations. In this study, forecasts from the high-resolution numerical weather prediction model AROME was used to analyze power production forecast performance for an offshore site in the Baltic Sea. To improve the AROME forecasts, six post-processing methods were investigated and their individual performance analyzed in general as well as for different wind speed ranges, boundary layer stratifications, synoptic situations and in low-level jet conditions. In general, AROME performed well in forecasting the power production, but applying smoothing or using a random forest algorithm increased forecast skill. Smoothing the forecast improved the performance at all wind speeds, all stratifications and for all synoptic weather classes, and the random forest method increased the forecast skill during low-level jets. To achieve the best performance, we recommend selecting which method to use based on the forecasted weather conditions. Combining forecasts from neighboring grid points, combining the recent forecast with the forecast from yesterday or applying linear regression to correct the forecast based on earlier performance were not fruitful methods to increase the overall forecast quality.

Bias-correcting input variables enhances forecasting of reference crop evapotranspiration

Reference crop evapotranspiration (ETo) is calculated using a standard formula with temperature, vapor pressure, solar radiation, and wind speed as input variables. ETo forecasts can be produced when forecasts of these input variables from numerical weather prediction (NWP) models are available. As raw ETo forecasts are often subject to systematic errors, statistical calibration is needed for improving forecast quality. The most straightforward and widely used approach is to directly calibrate raw ETo forecasts constructed with the raw forecasts of input variables. However, the predictable signal in ETo forecasts may not be fully implemented by this approach, which does not deal with error propagation from input variables to ETo forecasts. We hypothesize that correcting errors in input variables as a precursor to forecast calibration will lead to more skillful ETo forecasts. To test this hypothesis, we evaluate two calibration strategies that construct raw ETo forecasts with the raw (strategy i) or bias-corrected (strategy ii) input variables in ETo forecast calibration across Australia. Calibrated ETo forecasts based on bias-corrected input variables (strategy ii) demonstrate lower biases, higher correlation coefficients, and higher skills than forecasts produced by the calibration using raw input variables (strategy i). This investigation indicates that improving raw forecasts of input variables could effectively reduce error propagation and enhance ETo forecast calibration. We anticipate that future NWP-based ETo forecasting will benefit from adopting the calibration strategy developed in this study to produce more skillful ETo forecasts.