scholarly journals A Hybrid NWP–Analog Ensemble

2016 ◽  
Vol 144 (3) ◽  
pp. 897-911 ◽  
Author(s):  
F. Anthony Eckel ◽  
Luca Delle Monache
Keyword(s):  
Real Time ◽  
Weather Prediction ◽  
Ensemble Member ◽  
Initial Condition ◽  
Error Growth ◽  
High Quality ◽  
Time Model ◽  
Probabilistic Forecasts ◽  
Nwp Model ◽  
Dependent Error

Abstract An analog ensemble (AnEn) is constructed by first matching up the current forecast from a numerical weather prediction (NWP) model with similar past forecasts. The verifying observation from each match is then used as an ensemble member. For at least some applications, the advantages of AnEn over an NWP ensemble (multiple real-time model runs) may include higher efficiency, avoidance of initial condition and model perturbation challenges, and little or no need for postprocessing calibration. While AnEn can capture flow-dependent error growth, it may miss aspects of error growth that can be represented dynamically by the multiple real-time model runs of an NWP ensemble. To combine the strengths of the AnEn and NWP ensemble approaches, a hybrid ensemble (HyEn) is constructed by finding m analogs for each member of a small n-member NWP ensemble, to produce a total of m × n members. Forecast skill is compared between the AnEn, HyEn, and an NWP ensemble calibrated using logistic regression. The HyEn outperforms the other approaches for probabilistic 2-m temperature forecasts yet underperforms for 10-m wind speed. The mixed results reveal a dependence on the intrinsic skill of the NWP members employed. In this study, the NWP ensemble is underspread for both 2-m temperature and 10-m winds, yet displays some ability to represent flow-dependent error for the former and not the latter. Thus, the HyEn is a promising approach for efficient generation of high-quality probabilistic forecasts, but requires use of a small, and at least partially functional, NWP ensemble.

scholarly journals Statistical post-processing of wind speed forecasts using convolutional neural networks

2021 ◽  
Author(s):  
Simon Veldkamp ◽  
Kirien Whan ◽  
Sjoerd Dirksen ◽  
Maurice Schmeits
Keyword(s):  
Neural Networks ◽  
Wind Speed ◽  
Convolutional Neural Networks ◽  
Weather Prediction ◽  
Estimation Methods ◽  
Post Processing ◽  
Skill Scores ◽  
Probabilistic Forecasts ◽  
Truncated Normal ◽  
Nwp Model

AbstractCurrent statistical post-processing methods for probabilistic weather forecasting are not capable of using full spatial patterns from the numerical weather prediction (NWP) model. In this paper we incorporate spatial wind speed information by using convolutional neural networks (CNNs) and obtain probabilistic wind speed forecasts in the Netherlands for 48 hours ahead, based on KNMI’s deterministic Harmonie-Arome NWP model. The probabilistic forecasts from the CNNs are shown to have higher Brier skill scores for medium to higher wind speeds, as well as a better continuous ranked probability score (CRPS) and logarithmic score, than the forecasts from fully connected neural networks and quantile regression forests. As a secondary result, we have compared the CNNs using 3 different density estimation methods (quantized softmax (QS), kernel mixture networks, and fitting a truncated normal distribution), and found the probabilistic forecasts based on the QS method to be best.

scholarly journals Influence of Parameterization of Some Physical Processes in Soils on Numerical Meteorological Forecasts of Surface Fields

2016 ◽  
Vol 20 (4) ◽  
pp. 48-58
Author(s):  
Andrzej Mazur ◽  
Grzegorz Duniec
Keyword(s):  
Boundary Layer ◽  
Numerical Models ◽  
Weather Prediction ◽  
Water System ◽  
Soil Processes ◽  
Plant Water ◽  
Physical Processes ◽  
High Quality ◽  
Nwp Model ◽  
Surface Fields

Abstract Physical processes in soil-plant-water system are very complicated. Complex physical processes in soil, in particular interaction between soil-plant-water system have significant influence on processes in Planetary Boundary Layer. Changes of soil state can significantly modify processes in the PBL and meteorological fields. Since numerical models are to determine the forecast of high quality, the physical processes occurring in soil should be properly described and then appropriately introduced into a model. Every process in soil occurs on a smaller scale than original model’s domain, so it should be described via adequate parameterization. Overall, soil parameterizations implemented in current numerical weather prediction (NWP) model(s) were prepared almost 40 years ago, when NWP models worked with very poor resolution mesh. Since nowadays NWP works over domains of high resolution, these “old” schemes parameterization must be adequately revised. In this paper preliminary results of changes of parameterization of soil processes will be presented.

scholarly journals Analysis of Missed Summer Severe Rainfall Forecasts

2016 ◽  
Vol 31 (2) ◽  
pp. 433-450 ◽  
Author(s):  
Zuohao Cao ◽  
Da-Lin Zhang
Keyword(s):  
Convective Instability ◽  
Weather Prediction ◽  
Convective Available Potential Energy ◽  
Vertical Motion ◽  
Cloud Water ◽  
Initial Condition ◽  
Latent Heating ◽  
Initial Model ◽  
Training Process ◽  
Nwp Model

Abstract Despite considerable progress in mesoscale numerical weather prediction (NWP), the ability to predict summer severe rainfall (SSR) in terms of amount, location, and timing remains very limited because of its association with convective or mesoscale phenomena. In this study, two representative missed SSR events that occurred in the highly populated Great Lakes regions are analyzed within the context of moisture availability, convective instability, and lifting mechanism in order to help identify the possible causes of these events and improve SSR forecasts/nowcasts. Results reveal the following limitations of the Canadian regional NWP model in predicting SSR events: 1) the model-predicted rainfall is phase shifted to an undesired location that is likely caused by the model initial condition errors, and 2) the model is unable to resolve the echo-training process because of the weakness of the parameterized convection and/or coarse resolutions. These limitations are related to the ensuing model-predicted features: 1) vertical motion in the areas of SSR occurrence is unfavorable for triggering parameterized convection and grid-scale condensation; 2) convective available potential energy is lacking for initial model spinup and later for elevating latent heating to higher levels through parameterized convection, giving rise to less precipitation; and 3) the conversion of water vapor into cloud water at the upper and middle levels is underpredicted. Recommendations for future improvements are discussed.

scholarly journals Air Quality Predictions with an Analog Ensemble

2018 ◽  
Author(s):  
Luca Delle Monache ◽  
Stefano Alessandrini ◽  
Irina Djalalova ◽  
James Wilczak ◽  
Jason C. Knievel
Keyword(s):  
Air Quality ◽  
Environmental Protection ◽  
Real Time ◽  
Numerical Weather Prediction ◽  
Weather Prediction ◽  
Numerical Weather ◽  
Model Based ◽  
Computational Expense ◽  
Probabilistic Forecasts ◽  
A Current

Abstract. The authors demonstrate how the analog ensemble (AnEn) can efficiently generate deterministic and probabilistic forecasts of air quality. AnEn estimates the probability of future observations of a predictand based on a current deterministic numerical weather prediction and an archive of prior analog predictions paired with prior observations. The method avoids the complexity and real-time computational expense of dynamical (i.e., model-based) ensembles. The authors apply AnEn to observations from the Environmental Protection Agency's (EPA's) AIRNow network and to forecasts from the Community Multiscale Air Quality (CMAQ). Compared to raw forecasts from CMAQ, deterministic forecasts of O3 and PM2.5 based on AnEn's mean have lower errors, both systemic and random, and are better correlated with observations. Probabilistic forecasts from AnEn are statistically consistent, reliable, and sharp, and they quantify the uncertainty of the underlying prediction.

scholarly journals A methodology for probabilistic real-time forecasting – an urban case study

2012 ◽  
Vol 15 (3) ◽  
pp. 751-762 ◽  
Author(s):  
Jeanne-Rose René ◽  
Henrik Madsen ◽  
Ole Mark
Keyword(s):  
Real Time ◽  
Probability Distributions ◽  
Methodological Approach ◽  
Weather Prediction ◽  
Probabilistic Approach ◽  
Latin Hypercube Sampling ◽  
Social Consequences ◽  
Complex Nature ◽  
Modelling Uncertainty ◽  
Nwp Model

The phenomenon of urban flooding due to rainfall exceeding the design capacity of drainage systems is a global problem and can have significant economic and social consequences. The complex nature of quantitative precipitation forecasts (QPFs) from numerical weather prediction (NWP) models has facilitated a need to model and manage uncertainty. This paper presents a probabilistic approach for modelling uncertainty from single-valued QPFs at different forecast lead times. The uncertainty models in the form of probability distributions of rainfall forecasts combined with a sewer model is an important advancement in real-time forecasting at the urban scale. The methodological approach utilized in this paper involves a retrospective comparison between historical forecasted rainfall from a NWP model and observed rainfall from rain gauges from which conditional probability distributions of rainfall forecasts are derived. Two different sampling methods, respectively, a direct rainfall quantile approach and the Latin hypercube sampling-based method were used to determine the uncertainty in forecasted variables (water level, volume) for a test urban area, the city of Aarhus. The results show the potential for applying probabilistic rainfall forecasts and their subsequent use in urban drainage forecasting for estimation of prediction uncertainty.

scholarly journals Atmospheric forcing by ALADIN/MFSTEP and MFSTEP oriented tunings

2006 ◽  
Vol 3 (3) ◽  
pp. 319-342 ◽  
Author(s):  
R. Brožková ◽  
M. Derková ◽  
M. Belluš ◽  
F. Farda
Keyword(s):  
Real Time ◽  
Numerical Weather Prediction ◽  
Weather Forecasting ◽  
Weather Prediction ◽  
Fine Scale ◽  
Atmospheric Forcing ◽  
Numerical Weather ◽  
Real Time Mode ◽  
Nwp Model

Abstract. ALADIN/MFSTEP is a configuration of the numerical weather prediction (NWP) model ALADIN run in a dedicated real-time mode for the purposes of the MFSTEP Project. A special attention was paid to the quality of atmospheric fluxes used for the forcing of fine-scale oceanographic models. This paper describes the novelties applied in ALADIN/MFSTEP initiated by the MFSTEP demands, leading also to improvements in general weather forecasting.

Real-Time Water Vapor Maps from a GPS Surface Network: Construction, Validation, and Applications

2009 ◽  
Vol 48 (7) ◽  
pp. 1302-1316 ◽  
Author(s):  
Siebren de Haan ◽  
Iwan Holleman ◽  
Albert A. M. Holtslag
Keyword(s):  
Water Vapor ◽  
Standard Deviation ◽  
Real Time ◽  
Weather Prediction ◽  
Surface Wind ◽  
Step Method ◽  
Background Error ◽  
Surface Network ◽  
Nwp Model ◽  
Additional Value

Abstract In this paper the construction of real-time integrated water vapor (IWV) maps from a surface network of global positioning system (GPS) receivers is presented. The IWV maps are constructed using a two-dimensional variational technique with a persistence background that is 15 min old. The background error covariances are determined using a novel two-step method, which is based on the Hollingsworth–Lonnberg method. The quality of these maps is assessed by comparison with radiosonde observations and IWV maps from a numerical weather prediction (NWP) model. The analyzed GPS IWV maps have no bias against radiosonde observations and a small bias against NWP analysis and forecasts up to 9 h. The standard deviation with radiosonde observations is around 2 kg m−2, and the standard deviation with NWP increases with increasing forecast length (from 2 kg m−2 for the NWP analysis to 4 kg m−2 for a forecast length of 48 h). To illustrate the additional value of these real-time products for nowcasting, three thunderstorm cases are discussed. The constructed GPS IWV maps are combined with data from the weather radar, a lightning detection network, and surface wind observations. All cases show that the location of developing thunderstorms can be identified 2 h prior to initiation in the convergence of moist air.

scholarly journals The HMT Multi-Radar Multi-Sensor Hydro Experiment

2017 ◽  
Vol 98 (2) ◽  
pp. 347-359 ◽  
Author(s):  
Steven M. Martinaitis ◽  
Jonathan J. Gourley ◽  
Zachary L. Flamig ◽  
Elizabeth M. Argyle ◽  
Robert A. Clark ◽  
...  
Keyword(s):  
Real Time ◽  
Flash Flood ◽  
Weather Prediction ◽  
The United States ◽  
Test Bed ◽  
Flash Flooding ◽  
Spatial Coverage ◽  
Skill Scores ◽  
Probabilistic Forecasts ◽  
Flood Detection

Abstract There are numerous challenges with the forecasting and detection of flash floods, one of the deadliest weather phenomena in the United States. Statistical metrics of flash flood warnings over recent years depict a generally stagnant warning performance, while regional flash flood guidance utilized in warning operations was shown to have low skill scores. The Hydrometeorological Testbed—Hydrology (HMT-Hydro) experiment was created to allow operational forecasters to assess emerging products and techniques designed to improve the prediction and warning of flash flooding. Scientific goals of the HMT-Hydro experiment included the evaluation of gridded products from the Multi-Radar Multi-Sensor (MRMS) and Flooded Locations and Simulated Hydrographs (FLASH) product suites, including the experimental Coupled Routing and Excess Storage (CREST) model, the application of user-defined probabilistic forecasts in experimental flash flood watches and warnings, and the utility of the Hazard Services software interface with flash flood recommenders in real-time experimental warning operations. The HMT-Hydro experiment ran in collaboration with the Flash Flood and Intense Rainfall (FFaIR) experiment at the Weather Prediction Center to simulate the real-time workflow between a national center and a local forecast office, as well as to facilitate discussions on the challenges of short-term flash flood forecasting. Results from the HMT-Hydro experiment highlighted the utility of MRMS and FLASH products in identifying the spatial coverage and magnitude of flash flooding, while evaluating the perception and reliability of probabilistic forecasts in flash flood watches and warnings. NSSL scientists and NWS forecasters evaluate new tools and techniques through real-time test bed operations for the improvement of flash flood detection and warning operations.

scholarly journals A Comparative Verification of Localized Aviation Model Output Statistics Program (LAMP) and Numerical Weather Prediction (NWP) Model Forecasts of Ceiling Height and Visibility

2010 ◽  
Vol 25 (4) ◽  
pp. 1161-1178 ◽  
Author(s):  
David E. Rudack ◽  
Judy E. Ghirardelli
Keyword(s):  
United States ◽  
Short Range ◽  
Mesoscale Model ◽  
Weather Prediction ◽  
Weather Research And Forecasting ◽  
Model Output ◽  
Probabilistic Forecasts ◽  
Ceiling Height ◽  
Nwp Model ◽  
Model Output Statistics

Abstract In an effort to support aviation forecasting, the National Weather Service’s Meteorological Development Laboratory (MDL) has recently redeveloped the Localized Aviation Model Output Statistics (MOS) Program (LAMP) system. LAMP is designed to run hourly in NWS operations and produce short-range aviation forecast guidance at 1-h projections out to 25 h. This paper compares and contrasts LAMP ceiling height and visibility forecasts with forecasts produced by the 20-km Rapid Update Cycle model (RUC20), the Weather Research and Forecasting Nonhydrostatic Mesoscale Model (WRF-NMM), and the Short-Range Ensemble Forecast system (SREF). RUC20 and WRF-NMM forecasts of continuous ceiling height and visibility were interpolated to stations and converted into categorical forecasts. These interpolated forecasts were also categorized into instrument flight rule (IFR) or lower conditions and verified against LAMP forecasts at stations in the contiguous United States. LAMP and SREF probabilistic forecasts of ceiling height and visibility from LAMP and the SREF system were also verified. This study demonstrates that for the 0000 and 1200 UTC cycles over the contiguous United States, LAMP station-based categorical forecasts of ceiling height, visibility, and IFR conditions or lower are more accurate than the RUC20 and WRF-NMM ceiling height and visibility forecasts interpolated to stations. Moreover, for the 0900 and 2100 UTC forecast cycles and verification periods studied here, LAMP ceiling height and visibility probabilities exhibit better reliability and skill than the SREF system.

scholarly journals On the predictability of outliers in ensemble forecasts

2012 ◽  
Vol 8 (1) ◽  
pp. 53-57
Author(s):  
S. Siegert ◽  
J. Bröcker ◽  
H. Kantz
Keyword(s):  
Logistic Regression ◽  
Weather Prediction ◽  
Weather Conditions ◽  
Skill Score ◽  
Ensemble Forecast ◽  
Ensemble Member ◽  
Base Rate ◽  
Ensemble Forecasts ◽  
Probabilistic Forecasts ◽  
Ensemble Interpretation

Abstract. In numerical weather prediction, ensembles are used to retrieve probabilistic forecasts of future weather conditions. We consider events where the verification is smaller than the smallest, or larger than the largest ensemble member of a scalar ensemble forecast. These events are called outliers. In a statistically consistent K-member ensemble, outliers should occur with a base rate of 2/(K+1). In operational ensembles this base rate tends to be higher. We study the predictability of outlier events in terms of the Brier Skill Score and find that forecast probabilities can be calculated which are more skillful than the unconditional base rate. This is shown analytically for statistically consistent ensembles. Using logistic regression, forecast probabilities for outlier events in an operational ensemble are calculated. These probabilities exhibit positive skill which is quantitatively similar to the analytical results. Possible causes of these results as well as their consequences for ensemble interpretation are discussed.

Sign in / Sign up

Export Citation Format

Share Document