Numerical Weather Prediction Products Research Articles

Pre‐tactical convection prediction for air traffic flow management using LSTM neural network

AbstractThis paper aims to explore machine learning techniques for post‐processing high‐resolution Numerical Weather Prediction (NWP) products for the early detection of convection. Data from the Arome Ensemble Prediction System and satellite observations from the Rapidly Developing Thunderstorm (RDT) product by Météo‐France are used to train a recurrent neural network model to predict areas of total convection and moderate convection. The learning task is formulated as a binary classification problem using a long short‐term memory (LSTM) network architecture. Results from the LSTM model are compared with an object‐based probabilistic approach to forecast convection using metrics such as a receiver operating characteristics (ROC) curve, the Brier score and reliability. Results indicate that the LSTM model performs similarly to the object‐based probabilistic benchmark when classifying moderate convection areas and shows improved skill when classifying areas of total convective. Finally, the LSTM model results are presented within an air traffic management context to showcase the potential use of machine learning models within an operational application.

Spatial-temporal characteristics analysis of solar irradiance forecast errors in Europe and North America

Accurate photovoltaic (PV) power forecast is crucial to power systems. Solar irradiance forecast is the fundamentals of PV power forecast. Current researches merely focus on improving the forecast accuracy. There hasn't been comprehensive study on the spatial-temporal characteristics of solar irradiance forecast errors. Thus, this paper analyzed the error characteristics of solar irradiance forecast provided by European Centre for Medium-Range Weather Forecasts (ECMWF), one of the most accurate numerical weather prediction (NWP) products. ECMWF's forecasts were compared with ERA5 solar irradiance reanalysis data to reveal the error distribution characteristics. Four-year (2017–2020) data from the geographical region bounded by 63°N, −126°W, 21°S, and 36°E (covering most parts of Europe and North America) were studied. Experiments show that solar irradiance forecast errors peak at noon of the local time. Furthermore, correlations between solar irradiance forecast errors and other weather variables were also revealed. Experiments show that solar irradiance forecast errors have negative correlation with low cloud cover forecast errors and positive correlation with air temperature forecast errors. The possible reasons for the correlation relationship were also analyzed in detail. This paper systematically reveals the spatial-temporal characteristics of solar irradiance forecast errors and provides a useful guideline for solar PV system operation.

Evaluating the performance of WRF in simulating winds and surface meteorology during a Southern California wildfire event

The intensity and frequency of wildfires in California (CA) have increased in recent years, causing significant damage to human health and property. In October 2007, a number of small fire events, collectively referred to as the Witch Creek Fire or Witch Fire started in Southern CA and intensified under strong Santa Ana winds. As a test of current mesoscale modeling capabilities, we use the Weather Research and Forecasting (WRF) model to simulate the 2007 wildfire event in terms of meteorological conditions. The main objectives of the present study are to investigate the impact of horizontal grid resolution and planetary boundary layer (PBL) scheme on the model simulation of meteorological conditions associated with a Mega fire. We evaluate the predictive capability of the WRF model to simulate key meteorological and fire-weather forecast parameters such as wind, moisture, and temperature. Results of this study suggest that more accurate predictions of temperature and wind speed relevant for better prediction of wildfire spread can be achieved by downscaling regional numerical weather prediction products to 1 km resolution. Furthermore, accurate prediction of near-surface conditions depends on the choice of the planetary boundary layer parameterization. The MYNN parameterization yields more accurate prediction as compared to the YSU parameterization. WRF simulations at 1 km resolution result in better predictions of temperature and wind speed than relative humidity during the 2007 Witch Fire. In summary, the MYNN PBL parameterization scheme with finer grid resolution simulations improves the prediction of near-surface meteorological conditions during a wildfire event.

Evaluation of the Impact of Multi-Source Uncertainties on Meteorological and Hydrological Ensemble Forecasting

Evaluating the impact of multi-source uncertainties in complex forecasting systems is essential to understanding and improving the systems Previous studies have paid little attention to the influence of multi-source uncertainties in complex meteorological and hydrological forecasting systems. In this study, we developed a general ensemble framework based on Bayesian model averaging (BMA) for evaluating the impact of multi-source uncertainties in complex forecast systems. Based on this framework, we used eight numerical weather prediction products from the International Grand Global Ensemble (TIGGE) dataset, four hydrological models with different structures, and 1000 sets of parameters to comprehensively account for the input, structure, and parameter uncertainties. The framework’s application to the Chitan Basin in China revealed that the numerical weather prediction input uncertainty in the forecasting system was more significant than the hydrological model uncertainty. The hydrological model structure uncertainty was more prominent than the parameter uncertainty. The accuracy of the numerical weather prediction dominates the accuracy of the forecast of high flows. In addition, the structures and parameters of the hydrological model and their interactions contributed to the main uncertainty of the low flow forecasts. The streamflow was more realistically represented when the three uncertainty sources were considered jointly. By accounting for the significant uncertainty sources in complex forecast systems, the BMA ensemble forecasting produces more realistic and reliable predictions and reduces the influences of other incomplete considerations. The developed multi-source uncertainty assessment framework improves our understanding of the complex meteorological and hydrological forecasting system. Therefore, the framework is promising for improving the accuracy and reliability of complex forecasting systems.

Drought monitoring techniques for climate change in the Northern part of Thailand

Occurrence of drought is due to climate variability and climate change. Drought affected on agriculture, economic, social and environment. The suitable technique in detecting and monitoring drought events can provide useful information for prevention and mitigation planning from their impacts. The study objective was to investigate the change of rainfall and temperate as well as find out the appropriate drought index and its trend for the Northern part of Thailand. These drought indices were calculated by using meteorological data during 1951-2020 of 10 stations comparing with historical drought records between January and May during 2011-2016 in the same areas. The drought indices consisted of Standardized Precipitation Index (SPI), Deciles index and Moisture Available index (MAI) at 1-, 3- and 6-month timescales and Palmer Drought Severity Index (PDSI). It was found that PDSI was the most appropriate index. The Mann-Kendal test was used to investigate the trend of rainfall, temperature and PDSI. The results showed an increasing of temperature. There was no trend of drought intensity for long-term (40 years) periods but was in 4 stations for short-term (20 years) periods. The PDSI can be used for drought monitoring and predicting in advance by using numerical weather prediction products.

NWPP-EDH: Numerical weather prediction products evaporation duct regional prediction model

The evaporation duct is a special atmospheric stratification that can affect the propagation path of electromagnetic waves at sea, thus it is crucial for the ability of the radio communication systems. The traditional theoretical models of the evaporation duct often have limited accuracy. The actual observational data from voyages and stations are insufficient, and the existing data-driven evaporation duct height prediction models can only predict a particular point or route but cannot reproduce the regional distribution of the evaporation duct. To address these issues, we propose NWPP-EDH model in this work. The fitting ability of the NWPP-EDH model was tested. Its accuracy was compared with that of the Babin–Young–Carton model, Musson–Gauthier–Bruth model, and the classical Naval Postgraduate School model; compared to these models, the root mean square error (RMSE) of NWPP-EDH model was reduced by 71.8%, 87%, and 60.9%, respectively. Thus, we find that the model shows a better performance than the existing theoretical models.

Creating synthetic night-time visible-light meteorological satellite images using the GAN method

ABSTRACT Meteorology satellite visible-light images are critical for meteorologists. However, there are no satellite visible-light channels data at night, so we propose a method based on deep learning to create synthetic satellite visible-light images during night. Specifically, to produce realistic-looking products, we trained a generative adversarial network (GAN) model. The model can generate satellite visible-light images from corresponding satellite infrared (IR) channels data and numerical weather prediction (NWP) products. Considering to explicitly evaluating the contributions of different satellite IR channels and NWP products elements, we suggest using a channel-wise attention mechanic, e.g. a ‘Squeeze and Extraction Block’ (SEBlock) to quantitatively weigh the importance of different input data channels. The experiments based on the NWP products and the meteorology satellite data show that the proposed method is effective to create realistic synthetic satellite visible-light images during night.

Anomaly Based Synoptic Analysis and Model Prediction of Six Dust Storms Moving From Mongolia to Northern China in Spring 2021

AbstractSix dust storms that originated from Mongolia and moved to northern China in Spring 2021 are investigated by the anomaly variable‐based analysis method, which is compared to the traditional total‐variable based analysis method. Strong surface wind is a force to form dust storms as commonly known in current practice. However, dust area depicted by total‐wind model product analysis and operational forecast has a large bias to observation even for leading 24 hr. The analysis method of anomaly based variables showed that the domain of an airborne dust storm observed from satellite images is usually located between two neighboring convergence lines of 850 hPa anomalous winds. This connection of dust storm associated with the anomalous wind convergence extends southwestward from an anomalous cyclone and moves from Mongolia to northern China or Northeast China. It is thermodynamically associated with an anomalous thermal contrast, which has a spatial structure of wind, temperature, and geopotential height anomalies in the troposphere. We further found that numerical weather prediction products of the ECMWF model are able to indicate the anomalous wind pattern with lead times of 4–8 days by applying the anomaly based analysis approach. Therefore, this approach can not only depict dust area better but also reveal predictors much easier compared to the traditional total‐variable or full‐field‐based analysis method.

Observation and modelling of snow and sea ice mass balance and its sensitivity to atmospheric forcing during spring and summer 2007 in the Central

Snow depth and sea ice thickness were observed applying an ice mass balance buoy (IMB) in the drifting ice station Tara during the International Polar Year in 2007. Detailed in situ observations on meteorological variables and surface fluxes were taken during May to August. For this study, the operational analyses and short-term forecasts from two numerical weather prediction (NWP) models (ECMWF and HIRLAM) were extracted for the Tara drift trajectory. We compared the IMB, meteorological and surface flux observations against the NWP products, also applying a one-dimensional thermodynamic sea ice model (HIGHTSI) to calculate the snow and ice mass balance and its sensitivity to atmospheric forcing. The modelled snow depth time series, controlled by NWP-based precipitation, was in line with the observed one. HIGHTSI reproduced well the snowmelt onset, the progress of the melt, and the first date of snow-free conditions. HIGHTSI performed well also in the late August freezing season. Challenges remain to model the “false bottom” observed during the melting season. The evolution of the vertical temperature profiles in snow and ice was better simulated when the model was forced by in situ observations instead of NWP results. During the melting period, the nonlinear ice temperature profile was successfully modelled with both forcing options. During spring and the melting season, total sea ice mass balance was most sensitive to uncertainties in NWP results for the downward longwave radiation, followed by the downward shortwave radiation, air temperature, and wind speed. Citation : Cheng B, Vihma T, Palo T, et al. Observation and modelling of snow and sea ice mass balance and its sensitivity to atmospheric forcing during spring and summer 2007 in the Central Arctic. Adv Polar Sci, 2021, 32(4): 309-323, doi: 10.13679/j.advps.2021.0047

Wind Speed Forecasting in Fishing Harbor Anchorage Using a Novel Deep Convolutional Neural Network

Wind speed forecasting is an important issue in Marine fisheries. Improving the accuracy of wind speed forecasting is helpful to reduce the loss of fishery economy caused by strong wind. This paper proposes a wind speed forecasting method for fishing harbor anchorage based on a novel deep convolutional neural network. By combining the actual monitoring data of the automatic weather station with the numerical weather prediction (NWP) products, the proposed method constructing a deep convolutional neural network was based wind speed forecasting model. The model includes a one-dimensional convolution module (1D-CM) and a two-dimensional convolution module (2D-CM), in which 1D-CM extracts the time series features of the meteorological data, and 2D-CM is used to mine the latent semantic information from the outputs of 1D-CM. In order to alleviate the overfitting problem of the model, the L2 regularization and the dropout strategies are adopted in the training process, which improves the generalization of the model with higher reliability for wind speed prediction. Simulation experiments were carried out, using the 2016 wind speed and related meteorological data of a sheltered anchorage in Xiangshan, Ningbo, China. The results showed that, for wind speed forecast in the next 1 h, the proposed method outperform the traditional methods in terms of prediction accuracy; the mean absolute error (MAE) and the mean absolute percentage error (MAPE) of the proposed method are 0.3945 m/s and 5.71%, respectively.

Automatic Regional Interpretation and Forecasting System Supported by Machine Learning

The Model Output Statistics (MOS) model is a dynamic statistical weather forecast model based on multiple linear regression technology. It is greatly affected by the selection of parameters and predictors, especially when the weather changes drastically, or extreme weather occurs. We improved the traditional MOS model with the machine learning method to enhance the capabilities of self-learning and generalization. Simultaneously, multi-source meteorological data were used as the input to the model to improve the data quality. In the experiment, we selected the four areas of Nanjing, Beijing, Chengdu, and Guangzhou for verification, with the numerical weather prediction (NWP) products and observation data from automatic weather stations (AWSs) used to predict the temperature and wind speed in the next 24 h. From the experiment, it can be seen that the accuracy of the prediction values and speed of the method were improved by the ML-MOS. Finally, we compared the ML-MOS model with neural networks and support vector machine (SVM), the results show that the prediction result of the ML-MOS model is better than that of the above two models.

Convection indicator for pre-tactical air traffic flow management using neural networks

Convective weather is a large source of disruption for air traffic management operations. Being able to predict thunderstorms the day before operations can help traffic managers plan around weather and improve air traffic flow management operations. In this paper, machine learning is applied on data from satellite storm observations and ensemble numerical weather prediction products to detect convective weather 36 h in advance. The learning task is formulated as a binary classification problem and a neural network is trained to predict the occurrence of storms. The neural network results are used to develop a probabilistic based convection indicator capable of outperforming existing convection indicators found in the literature. Lastly, applications of the neural network based indicator in an air traffic management setting are presented.

Non-Rainy Attenuation Over Earth-Space Paths at Tropical and Temperate Sites Using Meteorological Data and NWP Products

A long-term investigation of the attenuation in non-rainy conditions has been carried out, for a tropical and a temperate location, using meteorological data and NWP (Numerical Weather Prediction) products during the period 2011-2015. The results show that ERA-5 full profiles are appropriate to estimate non-rainy attenuation in lieu of radiometric or radiosonde observations. Simpler regression-based methods are established. A new formulation for oxygen attenuation is introduced, which only requires surface temperature and pressure. Mass absorption coefficients are used for water vapour and cloud attenuation. Simpler regression-based approaches are then validated. The non-rainy attenuation at K, Ka and Q bands has been found noticeably higher in the tropics than in the temperate region. This study would facilitate the planning of global mobile satellite communication systems.

Which precipitation forecasts to use? Deterministic versus coarser‐resolution ensemble NWP models

AbstractDeterministic numerical weather prediction (NWP) models and ensemble NWP models are routinely run worldwide to assist weather forecasting. Deterministic forecasts are capable of capturing more detailed spatial features, while ensemble forecasts, often with a coarser resolution, have the ability to predict uncertainty in future conditions. A comparative understanding of the performance of these two types of forecasts is valuable for both users of NWP products and model developers. Past published comparisons tended to be limited in scope, for example, for only specific locations and weather events, and involving only raw forecasts. In this study, we conduct a comprehensive comparison of the performance of a deterministic model and an ensemble model of the Australian Bureau of Meteorology in forecasting daily precipitation across Australia over a period of 3 years. The deterministic model has a horizontal grid spacing of approximately 25 km, and the ensemble model 60 km. Despite the coarser resolution, the ensemble forecasts are found to be superior by a number of measures, including correlation, accuracy and reliability. This finding holds true for both raw forecasts from the NWP models and forecasts post‐processed using the recently developed seasonally coherent calibration (SCC) model. Post‐processing is shown to greatly improve the forecasts from both models; however, the improvement is greater for the deterministic model, narrowing the performance gap between the two models. This study adds strong evidence to the general notion that coarser‐resolution ensemble NWP forecasts perform better than deterministic forecasts.

Assessing forecasting performance of daily reference evapotranspiration using public weather forecast and numerical weather prediction

Accurate estimation of reference evapotranspiration (ET0) is important for water resource management and irrigation scheduling design. Limited by data availability of the numerical weather prediction (NWP), the public weather forecasting (PWF) has been widely used to forecast daily ET0 in China. Recently, two Chinese independently developed NWP products have been open to the public, namely, the Global/Regional Assimilation and Prediction System (GRAPES) and the T639 Global Medium-term Numerical Forecast Model (T639L60). In this paper, the forecast of ET0 is evaluated based on the recently released NWP outputs and the PWF using the FAO-56 PM equation. The daily ET0 forecast of the two datasets was compared against the ET0 that was calculated using weather variables observed from 31 automatic weather stations across a wide range of climate zones in China, including the temperate continental zone (TC), temperate monsoon zone (TM), mountain plateau zone (MP) and subtropical monsoon zone (SM). The results showed that the forecast performance of NWP for maximum and minimum air temperature (Tmax, Tmin) was worse due to the large errors in the altitude estimation process. The forecast performance of NWP for net radiation (Rnet) forecasted was better than that of PWF in the MP, TC and TM, while slightly worse in the SM. The VPD from the NWP was more accurate in TC and TM, while showed little difference with that from PWF in MP and SM. For the wind speed (u2), the NWP showed better forecast accuracy than PWF in all climate zones. Furthermore, the better forecast performance of ET0 was obtained by the NWP in the MP and TC. In the TM and SM, the PWF can provide a more accurate ET0 forecast compared to the NWP. Overall, the NWP can serve as a promising data source to generate acceptable ET0 forecasts across China. The high spatial resolution of the NWP provides the ability to capture the high resolution of the spatial variability in ET0, especially in the MP and TC where weather stations are sparse.

Optimizing Mooring Placement to Constrain Southern Ocean Air–Sea Fluxes

AbstractProposals from multiple nations to deploy air–sea flux moorings in the Southern Ocean have raised the question of how to optimize the placement of these moorings in order to maximize their utility, both as contributors to the network of observations assimilated in numerical weather prediction and also as a means to study a broad range of processes driving air–sea fluxes. This study, developed as a contribution to the Southern Ocean Observing System (SOOS), proposes criteria that can be used to determine mooring siting to obtain best estimates of net air–sea heat flux (Qnet). Flux moorings are envisioned as one component of a multiplatform observing system, providing valuable in situ point time series measurements to be used alongside satellite data and observations from autonomous platforms and ships. Assimilating models (e.g., numerical weather prediction and reanalysis products) then offer the ability to synthesize the observing system and map properties between observations. This paper develops a framework for designing mooring array configurations to maximize the independence and utility of observations. As a test case, within the meridional band from 35° to 65°S we select eight mooring sites optimized to explain the largest fraction of the total variance (and thus to ensure the least variance of residual components) in the area south of 20°S. Results yield different optimal mooring sites for low-frequency interannual heat fluxes compared with higher-frequency subseasonal fluxes. With eight moorings, we could explain a maximum of 24.6% of high-frequency Qnet variability or 44.7% of low-frequency Qnet variability.

Real‐Time Flood Forecasting Based on a High‐Performance 2‐D Hydrodynamic Model and Numerical Weather Predictions

AbstractA flood forecasting system commonly consists of at least two essential components, that is, a numerical weather prediction (NWP) model to provide rainfall forecasts and a hydrological/hydraulic model to predict the hydrological response. While being widely used for flood forecasting, hydrological models only provide a simplified representation of the physical processes of flooding due to negligence of strict momentum conservation. They cannot reliably predict the highly transient flooding process from intense rainfall, in which case a fully 2‐D hydrodynamic model is required. Due to high computational demand, hydrodynamic models have not been exploited to support real‐time flood forecasting across a large catchment at sufficiently high resolution. To fill the current research and practical gaps, this work develops a new forecasting system by coupling a graphics processing unit (GPU) accelerated hydrodynamic model with NWP products to provide high‐resolution, catchment‐scale forecasting of rainfall‐runoff and flooding processes induced by intense rainfall. The performance of this new forecasting system is tested and confirmed by applying it to “forecast” an extreme flood event across a 2,500‐km2 catchment at 10‐m resolution. Quantitative comparisons are made between the numerical predictions and field measurements in terms of water level and flood extent. To produce simulation results comparing well with the observations, the new flood forecasting system provides 34 hr of lead time when the weather forecasts are available 36 hr beforehand. Numerical experiments further confirm that uncertainties from the rainfall inputs are not amplified by the hydrodynamic model toward the final flood forecasting outputs in this case.

Evaluation of Precipitable Water Vapor from Five Reanalysis Products with Ground-Based GNSS Observations

At present, the global reliability and accuracy of Precipitable Water Vapor (PWV) from different reanalysis products have not been comprehensively evaluated. In this study, PWV values derived by 268 Global Navigation Satellite Systems (GNSS) stations around the world covering the period from 2016 to 2018 are used to evaluate the accuracies of PWV values from five reanalysis products. The temporal and spatial evolution is not taken into account in this analysis, although the temporal and spatial evolution of atmospheric flows is one of the most important information elements available in numerical weather prediction products. The evaluation results present that five reanalysis products with PWV accuracy from high to low are in the order of the fifth generation of European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis (ERA5), ERA-Interim, Japanese 55-year Reanalysis (JRA-55), National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR), and NCEP/DOE (Department of Energy) according to root mean square error (RMSE), bias and correlation coefficient. The ERA5 has the smallest RMSE value of 1.84 mm, while NCEP/NCAR and NCEP/DOE have bigger RMSE values of 3.34 mm and 3.51 mm, respectively. The findings demonstrate that ERA5 and two NCEP reanalysis products have the best and worst performance, respectively, among five reanalysis products. The differences in the accuracy of the five reanalysis products are mainly attributed to the differences in the spatial resolution of reanalysis products. There are some large absolute biases greater than 4 mm between GNSS PWV values and the PWV values of five reanalysis products in the southwest of South America and western China due to the limit of terrains and fewer observations. The accuracies of five reanalysis products are compared in different climatic zones. The results indicate that the absolute accuracies of five reanalysis products are highest in the polar regions and lowest in the tropics. Furthermore, the effects of different seasons on the accuracies of five reanalysis products are also analyzed, which indicates that RMSE values of five reanalysis products in summer and in winter are the largest and the smallest in the temperate regions. Evaluation results from five reanalysis products can help us to learn more about the advantages and disadvantages of the five released water vapor products and promote their applications.

Rainfall Estimation from Tropospheric Attenuation Affecting Satellite Links

A novel methodology for estimating rainfall rate from satellite signals is presented. The proposed inversion algorithm yields rain rate estimates by making opportunistic use of the downlink signal and exploiting local ancillary meteorological information (0 °C isotherm height and monthly convectivity index), which can be extracted on a Global basis from Numerical Weather Prediction (NWP) products. The methodology includes different expressions to take the different impact of stratiform and convective rain events on the link into due account. The model accuracy in predicting the rain rate is assessed (and compared to the one of other models), both on a statistical and on an instantaneous basis, by exploiting a full year of data collected in Milan, in the framework of the Alphasat Aldo Paraboni propagation experiment.

Methods to Estimate Gas Attenuation in Absence of a Radiometer to Support Satellite Propagation Experiments

The experimental characterization of tropospheric effects on electromagnetic waves relies on the reception of a beacon signal transmitted by a satellite. The total tropospheric attenuation can be derived from these measurements in combination with additional information, specifically the attenuation due to atmospheric gases, in turn derived from ancillary instruments or data sources. When available, this information is obtained from a cosited radiometer. This article presents three alternative procedures to estimate the gas attenuation that can be used in the absence of this instrument. The first one makes use of zenith total delay data obtained from the global navigation satellite system (GNSS) receivers and additional meteorological data. The second one relies on atmospheric profiles gathered from radiosonde observations and/or numerical weather prediction (NWP) products. The third one only makes use of standard atmospheric profiles. The accuracy of the procedures, when used in this application, is compared with the reference total tropospheric attenuation derived with the support of radiometric data, exploiting the Alphasat beacon measurements collected in Milan, Italy, in 2017. The results indicate that a better accuracy (average root mean square (rms) error below 0.1 dB when compared with the use of a radiometer) is achieved by using GNSS data because of their finer temporal resolution. Nonetheless, the three procedures can be equally recommended, their use being conditioned to the availability of the appropriate data in the area around the experimental site.