scholarly journals Comparing Area Probability Forecasts of (Extreme) Local Precipitation Using Parametric and Machine Learning Statistical Postprocessing Methods

2018 ◽  
Vol 146 (11) ◽  
pp. 3651-3673 ◽  
Author(s):  
Kirien Whan ◽  
Maurice Schmeits
Keyword(s):  
Machine Learning ◽  
Deterministic Model ◽  
Weather Prediction ◽  
Location Parameter ◽  
Important Predictor ◽  
Lead Times ◽  
Hourly Precipitation ◽  
Severe Thunderstorms ◽  
Probabilistic Forecasts ◽  
Precipitation Thresholds

Abstract Probabilistic forecasts, which communicate forecast uncertainties, enable users to make better weather-based decisions. Using precipitation and numerous instability indices from the deterministic model HARMONIE–AROME (HA; a nonhydrostatic numerical weather prediction model) as potential predictors, we generate summer areal probabilistic maximum hourly precipitation forecasts across 11 regions of the Netherlands. We compare the skill of three statistical postprocessing methods: an extended logistic regression (ELR), a zero-adjusted gamma distribution (ZAGA), and a machine learning-based method, quantile regression forests (QRF). Forecast skill for low and moderate precipitation thresholds increases with the inclusion of extra predictors, in addition to HA precipitation. HA precipitation is the most important predictor at all lead times in ELR and QRF, while in ZAGA, the most important predictor for the location parameter shifts over lead times from HA precipitation to indices of atmospheric instability. All three methods improve upon a climatological forecast for low and moderate precipitation thresholds. ZAGA and QRF are generally the most skillful methods at moderate thresholds. QRF tends to be the most skillful method at higher thresholds, particularly during the afternoon period. Forecasts are reliable at low and moderate thresholds but tend to be overconfident at higher thresholds. QRF and ZAGA have more potential economic value than the deterministic forecast, with value remaining at high thresholds. A maximum local hourly precipitation threshold of 30 mm h−1 (a criterion in the Royal Netherlands Meteorological Institute’s code yellow warning for severe thunderstorms) is skillfully forecast by QRF in the afternoon period at short lead times.

scholarly journals A framework for probabilistic weather forecast post-processing across models and lead times using machine learning

2021 ◽  
Vol 379 (2194) ◽  
pp. 20200099 ◽  
Author(s):  
Charlie Kirkwood ◽  
Theo Economou ◽  
Henry Odbert ◽  
Nicolas Pugeault
Keyword(s):  
Machine Learning ◽  
Decision Support ◽  
Probability Distributions ◽  
Weather Prediction ◽  
Weather Forecast ◽  
Predictive Distribution ◽  
Lead Times ◽  
Climate Modelling ◽  
Post Processing ◽  
Probabilistic Forecasts

Forecasting the weather is an increasingly data-intensive exercise. Numerical weather prediction (NWP) models are becoming more complex, with higher resolutions, and there are increasing numbers of different models in operation. While the forecasting skill of NWP models continues to improve, the number and complexity of these models poses a new challenge for the operational meteorologist: how should the information from all available models, each with their own unique biases and limitations, be combined in order to provide stakeholders with well-calibrated probabilistic forecasts to use in decision making? In this paper, we use a road surface temperature example to demonstrate a three-stage framework that uses machine learning to bridge the gap between sets of separate forecasts from NWP models and the ‘ideal’ forecast for decision support: probabilities of future weather outcomes. First, we use quantile regression forests to learn the error profile of each numerical model, and use these to apply empirically derived probability distributions to forecasts. Second, we combine these probabilistic forecasts using quantile averaging. Third, we interpolate between the aggregate quantiles in order to generate a full predictive distribution, which we demonstrate has properties suitable for decision support. Our results suggest that this approach provides an effective and operationally viable framework for the cohesive post-processing of weather forecasts across multiple models and lead times to produce a well-calibrated probabilistic output. This article is part of the theme issue ‘Machine learning for weather and climate modelling’.

Probabilistic Forecasts of (Severe) Thunderstorms for the Purpose of Issuing a Weather Alarm in the Netherlands

2008 ◽  
Vol 23 (6) ◽  
pp. 1253-1267 ◽  
Author(s):  
Maurice J. Schmeits ◽  
Kees J. Kok ◽  
Daan H. P. Vogelezang ◽  
Rudolf M. van Westrhenen
Keyword(s):  
The Netherlands ◽  
Weather Prediction ◽  
Limited Area ◽  
Regression Equations ◽  
Model Output ◽  
Forecast System ◽  
Severe Thunderstorms ◽  
Skill Scores ◽  
Probabilistic Forecasts ◽  
Total Lightning

Abstract The development and verification of a new model output statistics (MOS) system is described; this system is intended to help forecasters decide whether a weather alarm for severe thunderstorms, based on high total lightning intensity, should be issued in the Netherlands. The system consists of logistic regression equations for both the probability of thunderstorms and the conditional probability of severe thunderstorms in the warm half-year (from mid-April to mid-October). These equations have been derived for 12 regions of about 90 km × 80 km each and for projections out to 12 h in advance (with 6-h periods). As a source for the predictands, reprocessed total lightning data from the Surveillance et d’Alerte Foudre par Interférométrie Radioélectrique (SAFIR) network have been used. The potential predictor dataset not only consisted of the combined postprocessed output from two numerical weather prediction (NWP) models, as in previous work by the first three authors, but it also contained an ensemble of advected radar and lightning data for the 0–6-h projections. The NWP model output dataset contained 17 traditional thunderstorm indices, computed from a reforecasting experiment with the High-Resolution Limited-Area Model (HIRLAM) and postprocessed output from the European Centre for Medium-Range Weather Forecasts (ECMWF) model. Brier skill scores and attributes diagrams show that the skill of the MOS thunderstorm forecast system is good and that the severe thunderstorm forecast system generally is also skillful, compared to the 2000–04 climatology, and therefore, the preoperational system was made operational at the Royal Netherlands Meteorological Institute (KNMI) in 2008.

Using Artificial Neural Networks for Generating Probabilistic Subseasonal Precipitation Forecasts over California

2020 ◽  
Vol 148 (8) ◽  
pp. 3489-3506
Author(s):  
Michael Scheuerer ◽  
Matthew B. Switanek ◽  
Rochelle P. Worsnop ◽  
Thomas M. Hamill
Keyword(s):  
Neural Network ◽  
Large Scale ◽  
Signal To Noise Ratio ◽  
Weather Prediction ◽  
Forecast Skill ◽  
Lead Times ◽  
Probabilistic Forecasts ◽  
Medium Range ◽  
Artificial Neural ◽  
Artificial Neural Network Ann

Abstract Forecast skill of numerical weather prediction (NWP) models for precipitation accumulations over California is rather limited at subseasonal time scales, and the low signal-to-noise ratio makes it challenging to extract information that provides reliable probabilistic forecasts. A statistical postprocessing framework is proposed that uses an artificial neural network (ANN) to establish relationships between NWP ensemble forecast and gridded observed 7-day precipitation accumulations, and to model the increase or decrease of the probabilities for different precipitation categories relative to their climatological frequencies. Adding predictors with geographic information and location-specific normalization of forecast information permits the use of a single ANN for the entire forecast domain and thus reduces the risk of overfitting. In addition, a convolutional neural network (CNN) framework is proposed that extends the basic ANN and takes images of large-scale predictors as inputs that inform local increase or decrease of precipitation probabilities relative to climatology. Both methods are demonstrated with ECMWF ensemble reforecasts over California for lead times up to 4 weeks. They compare favorably with a state-of-the-art postprocessing technique developed for medium-range ensemble precipitation forecasts, and their forecast skill relative to climatology is positive everywhere within the domain. The magnitude of skill, however, is low for week-3 and week-4, and suggests that additional sources of predictability need to be explored.

scholarly journals Evaluation of the BMA probabilistic inflow forecasts using TIGGE numeric precipitation predictions based on artificial neural network

2018 ◽  
Vol 49 (5) ◽  
pp. 1417-1433 ◽  
Author(s):  
Yixuan Zhong ◽  
Shenglian Guo ◽  
Huanhuan Ba ◽  
Feng Xiong ◽  
Fi-John Chang ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Bayesian Model Averaging ◽  
Weather Prediction ◽  
Lead Times ◽  
Hydrologic Modelling ◽  
Probabilistic Forecasts ◽  
Inflow Forecasting ◽  
Artificial Neural ◽  
Reservoir Basin

Abstract Reservoir inflow forecasting is a crucial task for reservoir management. Without considering precipitation predictions, the lead time for inflow is subject to the concentration time of precipitation in the basin. With the development of numeric weather prediction (NWP) techniques, it is possible to forecast inflows with long lead times. Since larger uncertainty usually occurs during the forecasting process, much attention has been paid to probabilistic forecasts, which uses a probabilistic distribution function instead of a deterministic value to predict the future status. In this study, we aim at establishing a probabilistic inflow forecasting scheme in the Danjiangkou reservoir basin based on NWP data retrieved from the Interactive Grand Global Ensemble (TIGGE) database by using the Bayesian model averaging (BMA) method, and evaluating the skills of the probabilistic inflow forecasts. An artificial neural network (ANN) is used to implement hydrologic modelling. Results show that the corrected TIGGE NWP data can be applied sufficiently to inflow forecasting at 1–3 d lead times. Despite the fact that the raw ensemble inflow forecasts are unreliable, the BMA probabilistic inflow forecasts perform much better than the raw ensemble forecasts in terms of probabilistic style and deterministic style, indicating the established scheme can offer a useful approach to probabilistic inflow forecasting.

scholarly journals Rad-cGAN v1.0: Radar-based precipitation nowcasting model with conditional Generative Adversarial Networks for multiple domains

2021 ◽  
Author(s):  
Suyeon Choi ◽  
Yeonjoo Kim
Keyword(s):  
Machine Learning ◽  
South Korea ◽  
Transfer Learning ◽  
Weather Prediction ◽  
Generative Adversarial Networks ◽  
Lead Times ◽  
Successful Implementation ◽  
Short Term ◽  
Radar Images ◽  
Learning Technique

Abstract. Numerical weather prediction models and probabilistic extrapolation methods using radar images have been widely used for precipitation nowcasting. Recently, machine-learning-based precipitation nowcasting models have also been actively developed for relatively short-term precipitation predictions. This study aimed to develop a radar-based precipitation nowcasting model using an advanced machine learning technique, conditional generative adversarial network (cGAN), which shows high performance in image generation tasks. The cGAN-based precipitation nowcasting model, named Rad-cGAN, developed in this study was trained with a radar reflectivity map of the Soyang-gang Dam region in South Korea with a spatial domain of 128 × 128 km, spatial resolution of 1 km, and temporal resolution of 10 min. The model performance was evaluated using previously developed machine-learning-based precipitation nowcasting models, namely convolutional long short-term memory (ConvLSTM) and U-Net, as well as the baseline Eulerian persistence model. We demonstrated that Rad-cGAN outperformed other models not only for the chosen site but also for the entire domain across the Soyang-gang Dam region. Additionally, the proposed model maintained good performance even with lead times up to 80 min based on the critical success index at the intensity threshold of 0.1 mm h−1, while RainNet and ConvLSTM achieved lead times of 70 and 40 min, respectively. We also demonstrated the successful implementation of the transfer learning technique to efficiently train model with the data from other dam regions in South Korea, such as the Andong and Chungju Dam regions. We used pre-trained model, which was completely trained in the Soyang-gang Dam region. This study confirms that Rad-cGAN can be successfully applied to precipitation nowcasting with longer lead times, and using the transfer learning approach it shows good performance in regions other than the originally trained region.

scholarly journals A calibrated and consistent combination of probabilistic forecasts for the exceedance of several precipitation thresholds using neural networks.

2021 ◽  
Author(s):  
Peter Schaumann ◽  
Reinhold Hess ◽  
Martin Rempel ◽  
Ulrich Blahak ◽  
Volker Schmidt
Keyword(s):  
Neural Networks ◽  
Processing System ◽  
Skill Score ◽  
Horizontal Resolution ◽  
Grid Size ◽  
Ensemble Prediction ◽  
Lead Times ◽  
Ensemble Prediction System ◽  
Probabilistic Forecasts ◽  
Precipitation Thresholds

<p>In this talk we present a new statistical method for the seamless combination of two different ensemble precipitation forecasts (Nowcasting and NWP) using neural networks (NNs), see [1]. The method generates probabilistic forecasts for the exceedance of a set of predetermined thresholds (from 0.1mm up to 5mm). The aim of the combination model is to produce seamless and calibrated forecasts which outperform both input forecasts for all lead times and which are consistent regarding the considered thresholds. First, the hyper-parameters of the NNs are chosen according to a certain hyper-parameter optimization algorithm (not to be confused with the training of the NNs itself) on a 3-month dataset (dataset A). Then, the resulting NNs are tested via a rolling origin validation scheme on two 3-month datasets (datasets B & C) with different input forecasts each. Datasets A & B contain forecasts of DWD's RadVOR, a radar-based nowcasting system, and Ensemble-MOS, a post-processing system of NWP ensembles made by COSMO-DE-EPS, with a horizontal resolution of 20km, which is a predecessor of ICON-D2-EPS. Ensemble-MOS forecasts were provided for up to +6h, while RadVOR forecasts were available up to +2h. For dataset C, forecasts with a grid size of 2.2km are used from STEPS-DWD, a new implementation of the Short-term Ensemble Prediction System (STEPS) by  DWD, and ICON-D2-EPS as a NWP ensemble system. Forecasts were made up to +6h. In both validation datasets (B & C), the forecasts show the well-known behavior that the nowcasting systems RadVOR & STEPS are superior for short lead times, while NWP forecasts (Ensemble-MOS & ICON-D2-EPS) outperform these systems for later lead times. Based on the comparison of several validation scores (bias, Brier skill score, reliability and reliability diagram) we can show that the combination is indeed calibrated, consistent and outperforms both input forecasts for all lead times. It should be noted that the combination works on dataset C, although the hyper-parameters were chosen based on dataset A, which contains different forecasts for a different grid size.<br><br>[1] P. Schaumann, R. Hess, M. Rempel, U. Blahak and V. Schmidt, A calibrated and consistent combination of probabilistic forecasts for the exceedance of several precipitation thresholds using neural networks. Weather and Forecasting (in print)</p>

scholarly journals A calibrated and consistent combination of probabilistic forecasts for the exceedance of several precipitation thresholds using neural networks

2021 ◽  
Author(s):  
P. Schaumann ◽  
R. Hess ◽  
M. Rempel ◽  
U. Blahak ◽  
V. Schmidt
Keyword(s):  
Neural Networks ◽  
Weather Prediction ◽  
Short Term ◽  
Design Elements ◽  
Meaningful Information ◽  
Interaction Terms ◽  
Probabilistic Forecasts ◽  
Lower Accuracy ◽  
Precipitation Thresholds

AbstractThe seamless combination of nowcasting and numerical weather prediction (NWP) aims to provide a functional basis for very-short-term forecasts, that are essential e.g. for weather warnings. In this paper we propose a statistical method for precipitation using neural networks (NN) that combines nowcasting data from DWD’s radar based RadVOR system with post-processed forecasts of the high resolving NWP ensemble COSMO-DE-EPS. The postprocessing is performed by Ensemble-MOS of DWD. Whereas the quality of the nowcasting projections of RadVOR is excellent at the beginning, it declines rapidly after about 2 hours. The post-processed forecasts of COSMO-DE-EPS in contrast start with lower accuracy but provide meaningful information on longer forecast ranges. The combination of the two systems is performed for probabilities that the expected precipitation amounts exceed a series of predefined thresholds. The resulting probabilistic forecasts are calibrated and outperform both input systems in terms of accuracy for forecast ranges from 1 to 6 hours as shown by verification.The proposed NN-model generalises a previous statistical model based on extended logistic regression, which was restricted to only one threshold of 0.1 mm. The various layers of the NN-model are related to the conventional design elements (e.g. triangular functions and interaction terms) of the previous model for easier insight.

scholarly journals Low-visibility forecasts for different flight planning horizons using tree-based boosting models

2019 ◽  
Vol 5 (1) ◽  
pp. 101-114
Author(s):  
Sebastian J. Dietz ◽  
Philipp Kneringer ◽  
Georg J. Mayr ◽  
Achim Zeileis
Keyword(s):  
Traffic Management ◽  
Weather Prediction ◽  
Dew Point ◽  
Lead Times ◽  
Flight Planning ◽  
International Airport ◽  
Ordered Logistic Regression ◽  
Probabilistic Forecasts ◽  
Low Visibility ◽  
Nwp Model

Abstract. Low-visibility conditions enforce special procedures that reduce the operational flight capacity at airports. Accurate and probabilistic forecasts of these capacity-reducing low-visibility procedure (lvp) states help the air traffic management in optimizing flight planning and regulation. In this paper, we investigate nowcasts, medium-range forecasts, and the predictability limit of the lvp states at Vienna International Airport. The forecasts are generated with boosting trees, which outperform persistence, climatology, direct output of numerical weather prediction (NWP) models, and ordered logistic regression. The boosting trees consist of an ensemble of decision trees grown iteratively on information from previous trees. Their input is observations at Vienna International Airport as well as output of a high resolution and an ensemble NWP model. Observations have the highest impact for nowcasts up to a lead time of +2 h. Afterwards, a mix of observations and NWP forecast variables generates the most accurate predictions. With lead times longer than +7 h, NWP output dominates until the predictability limit is reached at +12 d. For lead times longer than +2 d, output from an ensemble of NWP models improves the forecast more than using a deterministic but finer resolved NWP model. The most important predictors for lead times up to +18 h are observations of lvp and dew point depression as well as NWP dew point depression. At longer lead times, dew point depression and evaporation from the NWP models are most important.

scholarly journals Statistical-Dynamical Forecasting of Sub-Seasonal North Atlantic Tropical Cyclone Occurrence

2021 ◽  
Author(s):  
Michael Maier-Gerber ◽  
Andreas H. Fink ◽  
Michael Riemer ◽  
Elmar Schoemer ◽  
Christoph Fischer ◽  
...  
Keyword(s):  
Tropical Cyclone ◽  
Gulf Of Mexico ◽  
North Atlantic ◽  
Statistical Models ◽  
Weather Prediction ◽  
Hybrid Approach ◽  
Lead Times ◽  
Probabilistic Forecasts ◽  
Tropical Conditions ◽  
Past Data

AbstractWhile previous research on sub-seasonal tropical cyclone (TC) occurrence has mostly focused on either the validation of numerical weather prediction (NWP) models, or the development of statistical models trained on past data, the present study combines both approaches to a statistical–dynamical model for probabilistic forecasts in the North Atlantic basin. Although state-of-the-art NWP models have been shown to lack predictive skill with respect to sub-seasonal weekly TC occurrence, they may predict the environmental conditions sufficiently well to generate predictors for a statistical model. Therefore, an extensive predictor set was generated, including predictor groups representing the climatological seasonal cycle (CSC), oceanic, and tropical conditions, tropical wave modes, as well as extratropical influences, respectively. The developed hybrid forecast model is systematically validated for the Gulf of Mexico and Central Main Development Region (MDR) for lead times up to five weeks. Moreover, its performance is compared against a statistical approach trained on past data, as well as against different climatological and NWP benchmarks. For sub-seasonal lead times, the CSC models are found to outperform the NWP models, which quickly loose skill within the first two forecast weeks, even in case of recalibration. The statistical models trained on past data increase skill over the CSC models, whereas even greater improvements in skill are gained by the hybrid approach out to week five. The vast majority of the additional sub-seasonal skill in the hybrid model, relative to the CSC model, could be attributed to the tropical (oceanic) conditions in the Gulf of Mexico (Central MDR).

Machine Learning Enhancement of Storm-Scale Ensemble Probabilistic Quantitative Precipitation Forecasts

2014 ◽  
Vol 29 (4) ◽  
pp. 1024-1043 ◽  
Author(s):  
David John Gagne ◽  
Amy McGovern ◽  
Ming Xue
Keyword(s):  
Machine Learning ◽  
Prediction Models ◽  
Weather Prediction ◽  
Precipitation Forecast ◽  
Spatial And Temporal Scales ◽  
Small Areas ◽  
Probabilistic Forecasts ◽  
Quantitative Precipitation Forecasts ◽  
Wide Variability ◽  
Added Uncertainty

Abstract Probabilistic quantitative precipitation forecasts challenge meteorologists due to the wide variability of precipitation amounts over small areas and their dependence on conditions at multiple spatial and temporal scales. Ensembles of convection-allowing numerical weather prediction models offer a way to produce improved precipitation forecasts and estimates of the forecast uncertainty. These models allow for the prediction of individual convective storms on the model grid, but they often displace the storms in space, time, and intensity, which results in added uncertainty. Machine learning methods can produce calibrated probabilistic forecasts from the raw ensemble data that correct for systemic biases in the ensemble precipitation forecast and incorporate additional uncertainty information from aggregations of the ensemble members and additional model variables. This study utilizes the 2010 Center for Analysis and Prediction of Storms Storm-Scale Ensemble Forecast system and the National Severe Storms Laboratory National Mosaic & Multi-Sensor Quantitative Precipitation Estimate as input data for training logistic regressions and random forests to produce a calibrated probabilistic quantitative precipitation forecast. The reliability and discrimination of the forecasts are compared through verification statistics and a case study.

Sign in / Sign up

Export Citation Format

Share Document