Numerical Weather Prediction Data Research Articles

Detectability of COVID-19 global emissions reductions in local CO2 concentration measurements

It is estimated that global anthropogenic carbon dioxide (CO2) emissions reduced by up to 12% at the start of 2020 compared to recent years due to the COVID-19 related downturn in economic activity. Despite the large decrease in CO2 emissions, no reduction in the trend in background atmospheric CO2 concentrations has been detected. So, how long would it take for sustained COVID-19 CO2 emission reductions to be detected in daily and monthly averaged local CO2 concentration measurements? CO2 concentration measurements for five measurement sites in the UK and Ireland are combined with meteorological numerical weather prediction data to build statistical models that can predict future CO2 concentrations. It is found that of the observed daily variability can be explained by these simple models. Emission reduction scenario experiments using these simple models illustrate that large daily and seasonal variability in local CO2 concentrations precludes the rapid emergence of a detectable signal. COVID-19 magnitude emissions reductions would only be detectable in the daily CO2 concentrations after at least 38 months and in monthly CO2 concentrations after 11 months of sustained reductions. For monthly CO2 concentrations the time of emergence is similar for all sites since the seasonal variability is largely driven by non-local fluxes of CO2 between the terrestrial biosphere and the atmosphere. The COVID-19 CO2 anthropogenic emissions reductions are similar in magnitude to those that are required to meet the Paris Agreement target of keeping global temperatures below C. This study demonstrates that, using measurements alone, there will be a considerable lag between changes in global anthropogenic emissions and a detected signal in local CO2 concentration trends. Thus, there is likely to be a delay of several years between changes in policy designed to meet CO2 anthropogenic emissions targets and our ability to detect the impact of these policies on CO2 concentrations using atmospheric measurements alone.

Feature extraction of meteorological factors for wind power prediction based on variable weight combined method

To achieve a high penetration of renewable energy integration, an effective solution is to explore the interdependence between numerical weather prediction (NWP) data and historical wind power to improve prediction accuracy. This paper proposes a novel combined approach for wind power prediction. The characteristics of NWP and historical wind power data are extracted by using the feature extraction technique, the predictor is designed based on extreme learning machine (ELM) and least squares support vector machine (LSSVM) model, and then key parameters of the prediction models are optimized by improving cuckoo search (ICS) to obtain a reliable value, which is defined as the pre-combined prediction value (PPA). To obtain a reliable result, a variance strategy is developed to allocate the weights of the pre-combined prediction model to obtain the final predicted values. Four seasons dataset collected from regional wind farms in China is utilized as a benchmark experiment to evaluate the effectiveness of the proposed approach. The results of comprehensive numerical cases with different seasons show that the proposed approach, which considers multiple-error metrics, including error metrics, accuracy rate, qualification rate, and improvement percentages, achieves higher accuracy than other benchmark prediction models.

Four‐Dimensional Wind Fields From Geostationary Hyperspectral Infrared Sounder Radiance Measurements With High Temporal Resolution

AbstractFour‐dimensional (4D) wind fields were derived from radiance measurements of the Geosynchronous Interferometric Infrared Sounder (GIIRS) onboard the FengYun‐4A satellite with 15‐min temporal resolution during Typhoon Maria (2018). Results are evaluated with independent ERA5 reanalysis, Global Data Assimilation System (GDAS) analysis and dropsonde wind profiles, and show a statistical root mean squared error less than 2 m/s for U and V components in troposphere against ERA5 and GDAS. The temporal variation of the wind fields from GIIRS at 15‐min intervals is consistent with that of the hourly ERA5. The added value of wind profiles over the numerical weather predictions (NWP) background field is also revealed. Further experiments confirm that higher temporal resolution from geostationary infrared (IR) sounder measurements could provide better dynamic information. 4D dynamic information can be extracted from high temporal resolution geostationary hyperspectral IR radiances in a consistent and continuous manner that can be used together with the thermodynamic information for various quantitative applications such as NWP data assimilation, near real‐time weather monitoring, situational awareness and nowcasting.

Local temperature forecasts based on statistical post‐processing of numerical weather prediction data

AbstractSix adaptive, short‐term post‐processing methods for correcting systematic errors in numerical weather prediction (NWP) forecasts of near‐surface air temperatures using local meteorological observations are assessed and compared. The methods tested are based on the simple moving average and the more advanced Kalman filter. Forecasts from the rather coarse‐resolution global NWP model Global Forecast System (GFS) and the regional high‐resolution NWP model HARMONIE are post‐processed, and the results are evaluated for 100 private weather stations in Denmark. The performance of the post‐processing methods differs depending on the NWP model. Overall, the combined moving average and a so‐called lead time Kalman filter performs the best. The moving average was shown to be superior to a diurnal bias correction Kalman filter at removing the longer‐term systematic errors for HARMONIE forecast data and comparable for GFS forecast data. Subsequent application of the lead time Kalman filter corrects for the short‐term errors using the real‐time forecast error. The post‐processing method is adaptive and there is no need for a long record of observations or a historical archive of forecasts.

Forecasting and uncertainty analysis of day-ahead photovoltaic power using a novel forecasting method

The primary means to promote grid-connected photovoltaic power generation is through accurately forecasting the power output from photovoltaic power stations. This paper proposes a method for day-ahead photovoltaic power forecasting (PPF) and uncertainty analysis using fuzzy c-means (FCM), whale optimization algorithm (WOA), least squares support vector machine (LSSVM), and non-parametric kernel density estimation (NPKDE). The FCM clustering algorithm was used to cluster historical data on numerical weather prediction and photovoltaic power stations, whereby daily data sharing similar meteorological information were clustered into one class. The rapid convergence speed and high convergence accuracy of the WOA were used to optimize the penalty factor and kernel function width of the LSSVM model; this was done to improve the calculation speed and forecasting accuracy of the LSSVM model. The WOA-LSSVM forecasting model was trained using the clustered numerical weather prediction and historical data of a photovoltaic power station. This was subsequently utilized to forecast day-ahead photovoltaic power. The NPKDE method was used to accurately calculate the probability density distribution of forecasting error and the confidence interval of the day-ahead PPF. The root mean square error (RMSE) values of the forecasting power of the WOA-LSSVM, PSO-LSSVM, LSSVM, LSTM and PSO-BP models are 2.55%, 3.00%, 5.60%, 6.03% and 3.18%, respectively, and the calculation results show that the forecasting accuracy of the WOA-LSSVM was higher relative to other models including PSO-LSSVM, LSSVM, LSTM and PSO-BP. Moreover, the NPKDE method was able to accurately describe the probability density distribution of the forecasting error.

A unified Bayesian filtering framework for multi-horizon wind speed prediction with improved accuracy

This paper proposes a unified filtering framework for multi-horizon wind speed prediction. The novelty of this paper focuses on the integration of the short-term prediction model, the Numerical Weather Prediction (NWP) and a smoothing term into a unified framework based on Bayesian filters. In the proposed framework, the system state function of the Bayesian filter is constructed by a pre-trained static model based on Gaussian Process Regression (GPR) to enhance the short-term prediction accuracy. Meanwhile, NWP data is integrated by the system input of the state function of the Bayesian filter. The integration of NWP guarantees the medium/long-term prediction accuracy. The measurement function of the Bayesian filter is constructed as a smoothing term to further improve the overall accuracy of the proposed method. The prediction accuracy of the proposed filtering framework is extensively benchmarked with other existing approaches based on the data from an offshore wind farm. The benchmarking results suggest that the proposed method yields improved prediction performance in short-term horizon. For medium/long-term horizon, the best accuracy of RMSE is improved by about 46% compared with the benchmarks.

A new gas absorption optical depth parameterisation for RTTOV version 13

Abstract. This paper describes a new gas optical depth parameterisation implemented in the most recent release, version 13, of the radiative transfer model RTTOV (Radiative Transfer for TOVS). RTTOV is a fast, one-dimensional radiative transfer model for simulating top-of-atmosphere visible, infrared, and microwave radiances observed by downward-viewing space-borne passive sensors. A key component of the model is the fast parameterisation of absorption by the various gases in the atmosphere. The existing parameterisation in RTTOV has been extended over many years to allow for additional variable gases in RTTOV simulations and to account for solar radiation and better support geostationary sensors by extending the validity to higher zenith angles. However, there are limitations inherent in the current approach which make it difficult to develop it further, for example by adding new variable gases. We describe a new parameterisation that can be applied across the whole spectrum, that allows for a wide range of zenith angles in support of solar radiation and geostationary sensors, and for which it will be easier to add new variable gases in support of user requirements. Comparisons against line-by-line radiative transfer simulations and against observations in the ECMWF operational system yield promising results, suggesting that the new parameterisation generally compares well with the old one in terms of accuracy. Further validation is planned, including testing in operational numerical weather prediction data assimilation systems.

Toward a Systematic Evaluation of Warm Conveyor Belts in Numerical Weather Prediction and Climate Models. Part I: Predictor Selection and Logistic Regression Model

AbstractThe physical and dynamical processes associated with warm conveyor belts (WCBs) importantly affect midlatitude dynamics and are sources of forecast uncertainty. Moreover, WCBs modulate the large-scale extratropical circulation and can communicate and amplify forecast errors. Therefore, it is desirable to assess the representation of WCBs in numerical weather prediction (NWP) models in particular on the medium to subseasonal forecast range. Most often, WCBs are identified as coherent bundles of Lagrangian trajectories that ascend in a time interval of 2 days from the lower to the upper troposphere. Although this Lagrangian approach has advanced the understanding of the involved processes significantly, the calculation of trajectories is computationally expensive and requires NWP data at a high spatial [], vertical [], and temporal resolution []. In this study, we present a statistical framework that derives footprints of WCBs from coarser NWP data that are routinely available. To this end, gridpoint-specific multivariate logistic regression models are developed for the Northern Hemisphere using meteorological parameters from ERA-Interim data as predictors and binary footprints of WCB inflow, ascent, and outflow based on a Lagrangian dataset as predictands. Stepwise forward selection identifies the most important predictors for these three WCB stages. The logistic models are reliable in replicating the climatological frequency of WCBs as well as the footprints of WCBs at instantaneous time steps. The novel framework is a first step toward a systematic evaluation of WCB representation in large datasets such as subseasonal ensemble reforecasts or climate projections.

Gridded Expansion of Forest Flux Observations and Mapping of Daily CO 2 Absorption by the Forests in Korea Using Numerical Weather Prediction Data and Satellite Images

Gridded Expansion of Forest Flux Observations and Mapping of Daily CO 2 Absorption by the Forests in Korea Using Numerical Weather Prediction Data and Satellite Images

Day‐ahead wind power combination forecasting based on corrected numerical weather prediction and entropy method

To satisfy the grid operation scheduling requirements for wind power forecasting model accuracy, the measured wind speed near the height of the wind turbine hub is added to the wind power combined forecasting model. First, the relationship between the numerical weather prediction wind speed and the measured wind speed at different heights are analysed, and the correlation between each wind speed and the wind power is compared. Second, the random forest algorithm combined with the cumulative contribution rate is used to select several meteorological types of numerical weather prediction data as the input of the long short-term memory network to predict wind speed. Third, while inputting the meteorological data provided by numerical weather prediction, which is highly related to wind power, the wind power prediction network also uses the predicted wind speed of the upper network as input to predict wind power. Finally, the entropy method is used to dynamically determine the combined weights of each forecasting model and improve the adaptability of the model. Research and analysis using measured data from two wind farms located in northeast China have verified the effectiveness of the method.

Validation of Aeolus winds using radiosonde observations and numerical weather prediction model equivalents

Abstract. In August 2018, the first Doppler wind lidar, developed by the European Space Agency (ESA), was launched on board the Aeolus satellite into space. Providing atmospheric wind profiles on a global basis, the Earth Explorer mission is expected to demonstrate improvements in the quality of numerical weather prediction (NWP). For the use of Aeolus observations in NWP data assimilation, a detailed characterization of the quality and the minimization of systematic errors is crucial. This study performs a statistical validation of Aeolus observations, using collocated radiosonde measurements and NWP forecast equivalents from two different global models, the ICOsahedral Nonhydrostatic model (ICON) of Deutscher Wetterdienst (DWD) and the European Centre for Medium-Range Weather Forecast (ECMWF) Integrated Forecast System (IFS) model, as reference data. For the time period from the satellite's launch to the end of December 2019, comparisons for the Northern Hemisphere (23.5–65∘ N) show strong variations of the Aeolus wind bias and differences between the ascending and descending orbit phase. The mean absolute bias for the selected validation area is found to be in the range of 1.8–2.3 m s−1 (Rayleigh) and 1.3–1.9 m s−1 (Mie), showing good agreement between the three independent reference data sets. Due to the greater representativeness errors associated with the comparisons using radiosonde observations, the random differences are larger for the validation with radiosondes compared to the model equivalent statistics. To achieve an estimate for the Aeolus instrumental error, the representativeness errors for the comparisons are determined, as well as the estimation of the model and radiosonde observational error. The resulting Aeolus error estimates are in the range of 4.1–4.4 m s−1 (Rayleigh) and 1.9–3.0 m s−1 (Mie). Investigations of the Rayleigh wind bias on a global scale show that in addition to the satellite flight direction and seasonal differences, the systematic differences vary with latitude. A latitude-based bias correction approach is able to reduce the bias, but a residual bias of 0.4–0.6 m s−1 with a temporal trend remains. Taking additional longitudinal differences into account, the bias can be reduced further by almost 50 %. Longitudinal variations are suggested to be linked to land–sea distribution and tropical convection that influences the thermal emission of the earth. Since 20 April 2020 a telescope temperature-based bias correction scheme has been applied operationally in the L2B processor, developed by the Aeolus Data Innovation and Science Cluster (DISC).

Continuous Multitrack Assimilation of Sentinel‐1 Precipitable Water Vapor Maps for Numerical Weather Prediction: How Far Can We Go With Current InSAR Data?

AbstractThe present study assesses the viability of including water vapor data from Interferometry Synthetic Aperture Radar (InSAR) in the initialization of numerical weather prediction (NWP) models, using already available Sentinel‐1 A and B products. Despite the limitations resulting from the 6‐day return period of images produced by the 2‐satellite system, it is found that for a sufficiently large domain designed to contain a set of images every 12 h (at varying locations), the impact on model performance is beneficial or at least neutral. The proposed methodology is tested in 24 consecutive 12 h forecasts, covering two cycles of the Sentinel‐1 system and 214 images, for a domain containing Iberia. A statistical analysis of the forecast precipitable water vapor (PWV) against independent GNSS observations concluded for relevant improvements in the different scores, especially during a consecutive 3‐day period where the standard initial data were less accurate. An analysis of the rain forecasts against gridded remote sensing observations further indicates an overall improvement in the grid‐point distribution of different precipitation classes throughout the simulation, even when the mean impact of PWV assimilation was not significant. It is suggested that current InSAR data are already a useful source of NWP data and will only become more relevant as new systems are put into operation.

A new simplified zenith tropospheric delay model for real-time GNSS applications

We have proposed a new simplified zenith tropospheric delay model for Global Navigation Satellite System (GNSS) real-time applications using 10 years of numerical weather prediction data from the National Center for Environmental Prediction (NCEP). Our proposed model is a blind model in which the atmospheric input parameters are tabulated based on realistic atmospheric behavior. The model with only 108 meteorological parameters is as simple as UNB3m model but has better performance; thus, it has great potential to be widely hardwired in GNSS receivers. Additionally, the accuracy of the new model is better than that of the state-of-the-art model named GPT2w in terms of the zenith tropospheric hydrostatic delay (ZHD) prediction, particularly at high altitudes. The proposed model is validated using NCEP reanalysis data and GNSS-derived International GNSS Service data. In comparison with the NCEP reanalysis data, the new model demonstrates an approximately 23% improvement in the root mean square (RMS) and a 76% improvement in the bias compared with the UNB3m model at the surface level. More significantly, this model performs better than the GPT2w model with respect to the ZHD prediction with increasing altitude, achieving up to a 96% improvement at an altitude of 10 km. The results of the zenith tropospheric delay (ZTD) model are compared with those derived from GNSS data. Notably, the RMS error and bias of the proposed model (with average values of 4.2 cm and 0.3 cm) are lower than those of the UNB3m model (with values of 5.0 cm and 1.0 cm), respectively.

Estimation of Ground-level PM 10 and PM 2.5 Concentrations Using Boosting-based Machine Learning from Satellite and Numerical Weather Prediction Data

Estimation of Ground-level PM 10 and PM 2.5 Concentrations Using Boosting-based Machine Learning from Satellite and Numerical Weather Prediction Data

Implementation of Spatial Downscaling Method Based on Gradient and Inverse Distance Squared (GIDS) for High-Resolution Numerical Weather Prediction Data

Implementation of Spatial Downscaling Method Based on Gradient and Inverse Distance Squared (GIDS) for High-Resolution Numerical Weather Prediction Data

Probabilistic Forecasting of Wind Turbine Icing Related Production Losses Using Quantile Regression Forests

A probabilistic machine learning method is applied to icing related production loss forecasts for wind energy in cold climates. The employed method, called quantile regression forests, is based on the random forest regression algorithm. Based on the performed tests on data from four Swedish wind parks available for two winter seasons, it has been shown to produce valuable probabilistic forecasts. Even with the limited amount of training and test data that were used in the study, the estimated forecast uncertainty adds more value to the forecast when compared to a deterministic forecast and a previously published probabilistic forecast method. It is also shown that the output from a physical icing model provides useful information to the machine learning method, as its usage results in an increased forecast skill when compared to only using Numerical Weather Prediction data. A potential additional benefit in machine learning for some stations was also found when using information in the training from other stations that are also affected by icing. This increases the amount of data, which is otherwise a challenge when developing forecasting methods for wind energy in cold climates.

Photo-voltaic power intra-day and daily statistical predictions using sum models composed from L-transformed PDE components in nodes of step by step developed polynomial neural networks

Precise forecasts of photo-voltaic (PV) energy production are necessary for its planning, utilization and integration into the electrical grid. Intra-day or daily statistical models, using only the latest weather observations and power data measurements, can predict PV power for a plant-specific location and condition on time. Numerical weather prediction (NWP) systems are run every 6 h to produce free prognoses of local cloudiness with a considerable delay and usually not in operational quality. Differential polynomial neural network (D-PNN) is a novel neuro-computing technique able to model complex weather patterns. D-PNN decomposes the n-variable partial differential equation (PDE), allowing complex representation of the near-ground atmospheric dynamics, into a set of 2-input node sub-PDEs. These are converted and substituted using the Laplace transformation according to operational calculus. D-PNN produces applicable PDE components which extend, one by one, its composite models using the selected optimal inputs. The models are developed with historical spatial data from estimated daily training periods for a specific inputs- > output time-shift to predict clear-sky index. Multi-step 1–9 h and one-step 24-h PV power predictions using machine learning and regression are compared to assess the performance of their models for both of the approaches. The presented spatial models obtain a better prediction accuracy than those post-processing NWP data, using a few variables only. The daily statistical models allow prediction of full PVP cycles in one step with an adequate accuracy in the morning and afternoon hours. This is inevitable in management of PV plant energy production and consumption.

Extensive comparison of physical models for photovoltaic power forecasting

Forecasting the power production of grid-connected photovoltaic (PV) power plants is essential for both the profitability and the prospects of the technology. Physically inspired modelling represents a common approach in calculating the expected power output from numerical weather prediction data. The model selection has a high effect on physical PV power forecasting accuracy, as the difference between the most and least accurate model chains is 13% in mean absolute error (MAE), 12% in root mean square error (RMSE), and 23–33% in skill scores for a PV plant on average. The power forecast performance analysis performed and verified for one-year 15-min resolution production data of 16 PV plants in Hungary for day-ahead and intraday time horizons on all possible combinations of nine direct and diffuse irradiance separation, ten tilted irradiance transposition, three reflection loss, five cell temperature, four PV module performance, two shading loss, and three inverter models.The two most critical calculation steps are identified as irradiance separation and transposition modelling, while the inverter models are the least important. Absolute and squared errors are two conflicting metrics, as the more detailed models result in the lowest MAE, while the simplest ones have the lowest RMSE. Wind speed forecasts have only a marginal effect on the PV power prediction. The results of this study contribute to a deeper understanding of the physical forecasting approach in the research community, while the main conclusions are also beneficial for PV plant owners in preparing their generation forecasts.

Retrieval of Daily Reference Evapotranspiration for Croplands in South Korea Using Machine Learning with Satellite Images and Numerical Weather Prediction Data

Evapotranspiration (ET) is an important component of the Earth’s energy and water cycle via the interaction between the atmosphere and the land surface. The reference evapotranspiration (ET0) is particularly important in the croplands because it is a convenient and reasonable method for calculating the actual evapotranspiration (AET) that represents the loss of water in the croplands through the soil evaporation and vegetation transpiration. To date, many efforts have been made to retrieve ET0 on a spatially continuous grid. In particular, the Moderate Resolution Imaging Spectroradiometer (MODIS) product is provided with a reasonable spatial resolution of 500 m and a temporal resolution of 8 days. However, the applicability to the local-scale variabilities due to complex and heterogeneous land surfaces in countries like South Korea is not sufficiently validated. Meanwhile, the AI approaches showed a useful functionality for the ET0 retrieval on the local scale but have rarely demonstrated a substantial product for a spatially continuous grid. This paper presented a retrieval of the daily reference evapotranspiration (ET0) over a 500 m grid for croplands in South Korea using machine learning (ML) with satellite images and numerical weather prediction data. In a blind test for 2013–2019, the ML-based ET0 model produced the accuracy statistics with a root mean square error of 1.038 mm/day and a correlation coefficient of 0.870. The results of the blind test were stable irrespective of location, year, and month. This outcome is presumably because the input data of the ML-based ET0 model were suitably arranged spatially and temporally, and the optimization of the model was appropriate. We found that the relative humidity and land surface temperature were the most influential variables for the ML-based ET0 model, but the variables with lower importance were also necessary to consider the nonlinearity between the variables. Using the daily ET0 data produced over the 500 m grid, we conducted a case study to examine agrometeorological characteristics of the croplands in South Korea during the period when heatwave and drought events occurred. Through the experiments, the feasibility of the ML-based ET0 retrieval was validated, especially for local agrometeorological applications in regions with heterogeneous land surfaces, such as South Korea.

Comparison of Machine-Learning Algorithms for Near-Surface Air-Temperature Estimation from FY-4A AGRI Data

Six machine-learning approaches, including multivariate linear regression (MLR), gradient boosting decision tree, k-nearest neighbors, random forest, extreme gradient boosting (XGB), and deep neural network (DNN), were compared for near-surface air-temperature (Tair) estimation from the new generation of Chinese geostationary meteorological satellite Fengyun-4A (FY-4A) observations. The brightness temperatures in split-window channels from the Advanced Geostationary Radiation Imager (AGRI) of FY-4A and numerical weather prediction data from the global forecast system were used as the predictor variables for Tair estimation. The performance of each model and the temporal and spatial distribution of the estimated Tair errors were analyzed. The results showed that the XGB model had better overall performance, with R2 of 0.902, bias of −0.087°C, and root-mean-square error of 1.946°C. The spatial variation characteristics of the Tair error of the XGB method were less obvious than those of the other methods. The XGB model can provide more stable and high-precision Tair for a large-scale Tair estimation over China and can serve as a reference for Tair estimation based on machine-learning models.