The Experts below are selected from a list of 11004 Experts worldwide ranked by ideXlab platform
Michael Scheuerer - One of the best experts on this subject based on the ideXlab platform.
-
probabilistic quantitative precipitation forecasting using ensemble model output statistics
Quarterly Journal of the Royal Meteorological Society, 2014Co-Authors: Michael ScheuererAbstract:Statistical post-processing of dynamical forecast ensembles is an essential component of weather forecasting. In this article, we present a post-processing method which generates full predictive probability distributions for precipitation accumulations based on ensemble model output statistics (EMOS). We model precipitation amounts by a generalized extreme value distribution which is left-censored at zero. This distribution permits modelling precipitation on the original scale without prior transformation of the data. A closed form expression for its continuous ranked probability score can be derived and permits computationally efficient model fitting. We discuss an extension of our approach which incorporates further statistics characterizing the spatial variability of precipitation amounts in the vicinity of the location of interest. The proposed EMOS method is applied to daily 18 h forecasts of 6 h accumulated precipitation over Germany in 2011 using the COSMO-DE ensemble prediction system operated by the German Meteorological Service. It yields calibrated and sharp predictive distributions and compares favourably with extended logistic regression and Bayesian model averaging which are state-of-the-art approaches for precipitation post-processing. The incorporation of neighbourhood information further improves predictive performance and turns out to be a useful strategy to account for displacement errors of the dynamical forecasts in a probabilistic forecasting framework.
-
probabilistic quantitative precipitation forecasting using ensemble model output statistics
arXiv: Applications, 2013Co-Authors: Michael ScheuererAbstract:Statistical post-processing of dynamical forecast ensembles is an essential component of weather forecasting. In this article, we present a post-processing method that generates full predictive probability distributions for precipitation accumulations based on ensemble model output statistics (EMOS). We model precipitation amounts by a generalized extreme value distribution that is left-censored at zero. This distribution permits modelling precipitation on the original scale without prior transformation of the data. A closed form expression for its continuous rank probability score can be derived and permits computationally efficient model fitting. We discuss an extension of our approach that incorporates further statistics characterizing the spatial variability of precipitation amounts in the vicinity of the location of interest. The proposed EMOS method is applied to daily 18-h forecasts of 6-h accumulated precipitation over Germany in 2011 using the COSMO-DE ensemble prediction system operated by the German Meteorological Service. It yields calibrated and sharp predictive distributions and compares favourably with extended logistic regression and Bayesian model averaging which are state of the art approaches for precipitation post-processing. The incorporation of neighbourhood information further improves predictive performance and turns out to be a useful strategy to account for displacement errors of the dynamical forecasts in a probabilistic forecasting framework.
Sebastian Lerch - One of the best experts on this subject based on the ideXlab platform.
-
log normal distribution based ensemble model output statistics models for probabilistic wind speed forecasting
Quarterly Journal of the Royal Meteorological Society, 2015Co-Authors: Sandor Baran, Sebastian LerchAbstract:Ensembles of forecasts are obtained from multiple runs of numerical weather forecasting models with different initial conditions and typically employed to account for forecast uncertainties. However, biases and dispersion errors often occur in forecast ensembles: they are usually underdispersive and uncalibrated and require statistical post-processing. We present an Ensemble Model Output Statistics (EMOS) method for calibration of wind-speed forecasts based on the log-normal (LN) distribution and we also show a regime-switching extension of the model, which combines the previously studied truncated normal (TN) distribution with the LN. Both models are applied to wind-speed forecasts of the eight-member University of Washington mesoscale ensemble, the 50 member European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble and the 11 member Aire Limitee Adaptation dynamique Developpement International-Hungary Ensemble Prediction System (ALADIN-HUNEPS) ensemble of the Hungarian Meteorological Service; their predictive performance is compared with that of the TN and general extreme value (GEV) distribution based EMOS methods and the TN–GEV mixture model. The results indicate improved calibration of probabilistic forecasts and accuracy of point forecasts in comparison with the raw ensemble and climatological forecasts. Further, the TN–LN mixture model outperforms the traditional TN method and its predictive performance is able to keep up with models utilizing the GEV distribution without assigning mass to negative values.
-
log normal distribution based emos models for probabilistic wind speed forecasting
arXiv: Methodology, 2014Co-Authors: Sandor Baran, Sebastian LerchAbstract:Ensembles of forecasts are obtained from multiple runs of numerical weather forecasting models with different initial conditions and typically employed to account for forecast uncertainties. However, biases and dispersion errors often occur in forecast ensembles, they are usually under-dispersive and uncalibrated and require statistical post-processing. We present an Ensemble Model Output Statistics (EMOS) method for calibration of wind speed forecasts based on the log-normal (LN) distribution, and we also show a regime-switching extension of the model which combines the previously studied truncated normal (TN) distribution with the LN. Both presented models are applied to wind speed forecasts of the eight-member University of Washington mesoscale ensemble, of the fifty-member ECMWF ensemble and of the eleven-member ALADIN-HUNEPS ensemble of the Hungarian Meteorological Service, and their predictive performances are compared to those of the TN and general extreme value (GEV) distribution based EMOS methods and to the TN-GEV mixture model. The results indicate improved calibration of probabilistic and accuracy of point forecasts in comparison to the raw ensemble and to climatological forecasts. Further, the TN-LN mixture model outperforms the traditional TN method and its predictive performance is able to keep up with the models utilizing the GEV distribution without assigning mass to negative values.
Sandor Baran - One of the best experts on this subject based on the ideXlab platform.
-
log normal distribution based ensemble model output statistics models for probabilistic wind speed forecasting
Quarterly Journal of the Royal Meteorological Society, 2015Co-Authors: Sandor Baran, Sebastian LerchAbstract:Ensembles of forecasts are obtained from multiple runs of numerical weather forecasting models with different initial conditions and typically employed to account for forecast uncertainties. However, biases and dispersion errors often occur in forecast ensembles: they are usually underdispersive and uncalibrated and require statistical post-processing. We present an Ensemble Model Output Statistics (EMOS) method for calibration of wind-speed forecasts based on the log-normal (LN) distribution and we also show a regime-switching extension of the model, which combines the previously studied truncated normal (TN) distribution with the LN. Both models are applied to wind-speed forecasts of the eight-member University of Washington mesoscale ensemble, the 50 member European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble and the 11 member Aire Limitee Adaptation dynamique Developpement International-Hungary Ensemble Prediction System (ALADIN-HUNEPS) ensemble of the Hungarian Meteorological Service; their predictive performance is compared with that of the TN and general extreme value (GEV) distribution based EMOS methods and the TN–GEV mixture model. The results indicate improved calibration of probabilistic forecasts and accuracy of point forecasts in comparison with the raw ensemble and climatological forecasts. Further, the TN–LN mixture model outperforms the traditional TN method and its predictive performance is able to keep up with models utilizing the GEV distribution without assigning mass to negative values.
-
log normal distribution based emos models for probabilistic wind speed forecasting
arXiv: Methodology, 2014Co-Authors: Sandor Baran, Sebastian LerchAbstract:Ensembles of forecasts are obtained from multiple runs of numerical weather forecasting models with different initial conditions and typically employed to account for forecast uncertainties. However, biases and dispersion errors often occur in forecast ensembles, they are usually under-dispersive and uncalibrated and require statistical post-processing. We present an Ensemble Model Output Statistics (EMOS) method for calibration of wind speed forecasts based on the log-normal (LN) distribution, and we also show a regime-switching extension of the model which combines the previously studied truncated normal (TN) distribution with the LN. Both presented models are applied to wind speed forecasts of the eight-member University of Washington mesoscale ensemble, of the fifty-member ECMWF ensemble and of the eleven-member ALADIN-HUNEPS ensemble of the Hungarian Meteorological Service, and their predictive performances are compared to those of the TN and general extreme value (GEV) distribution based EMOS methods and to the TN-GEV mixture model. The results indicate improved calibration of probabilistic and accuracy of point forecasts in comparison to the raw ensemble and to climatological forecasts. Further, the TN-LN mixture model outperforms the traditional TN method and its predictive performance is able to keep up with the models utilizing the GEV distribution without assigning mass to negative values.
-
statistical post processing of probabilistic wind speed forecasting in hungary
Meteorologische Zeitschrift, 2013Co-Authors: Sandor Baran, Andras Horanyi, Dora NemodaAbstract:Prediction of various weather quantities is mostly based on deterministic numerical weather forecasting models. Multiple runs of these models with different initial conditions result ensembles of forecasts which are applied for estimating the distribution of future weather quantities. However, the ensembles are usually under-dispersive and uncalibrated, so post-processing is required. In the present work Bayesian Model Averaging (BMA) is applied for calibrating ensembles of wind speed forecasts produced by the operational Limited Area Model Ensemble Prediction System of the Hungarian Meteorological Service (HMS). We describe two possible BMA models for wind speed data of the HMS and show that BMA post-processing significantly improves the calibration and precision of forecasts.
Joan Bech - One of the best experts on this subject based on the ideXlab platform.
-
a study of a heavy rainfall event and a tornado outbreak during the passage of a squall line over catalonia
Atmospheric Research, 2009Co-Authors: Jordi Mateo, Dolors Ballart, Clara Brucet, Montserrat Aran, Joan BechAbstract:Abstract During the 13 September 2006 a squall line crossed the coastal areas of Catalonia (NE of Spain) causing heavy rainfall and a tornado outbreak. In the south and central coast of Catalonia some waterspouts and tornadoes were observed. There was damage on roofs, walls and cars. Some cars were observed totally overturned and at least two trucks were diverted from their way. A description of the event is presented in this work. The synoptic and mesoscale patterns and possible tornadogenetic mechanisms are analyzed. The aim of this work is to document in detail this squall line event and to improve our understanding of convergence lines in this area. In the study satellite and radar images, rawinsonde data and automatic weather station observations of the Catalan Meteorological Service were examined to document and discuss the main features of this event.
-
uncertainty of precipitation estimates in convective events by the Meteorological Service of catalonia radar network
Atmospheric Research, 2009Co-Authors: Laura Trapero, Joan Bech, Nicolau Pineda, Tomeu Rigo, David ForcadellAbstract:Abstract In order to quantify the uncertainty of the radar-derived surface point quantitative precipitation estimates (QPE) from a regional radar network, a comparison has been made with a network of rain gauges. Three C-band Doppler radars and 161 telemetered gauges have been used. Both networks cover the area of Catalonia (NE Spain). Hourly accumulations integrated in daily amounts are studied. For each radar, three different precipitation products are obtained: short range, long range, and short range corrected radar QPE. The corrected product is generated by the HydroMeteorological Integrated Forecasting Tool (EHIMI), a software package designed to correct radar observations in real time for its use in hydroMeteorological applications. Among other features, EHIMI includes a topographical beam blockage correction procedure. The first part of the analysis examines the bias found in the radar. The three radars generally underestimate precipitation, an effect increased with range from the radar and beam blockage, which is examined in detail in this study. Moreover, corrected QPEs systematically improve the BIAS (2 dB) and RMSf for high blockages (50–70%). The second part of the analysis illustrates the temporal evolution of the daily mean bias. Finally, the uncertainty of each rain gauge has been compared to each rainfall radar product. Geographic distribution of daily BIAS is consistent with slight under-estimation at short range and substantial at long range, especially in the north of Catalonia, which is an area with important beam blockage (> 40%). These results contribute to improve the knowledge about the spatial distribution of the QPE error benefiting a number of applications including verification of high-resolution NWP precipitation forecasts and use of advanced hydroMeteorological models.
-
the weather radar network of the catalan Meteorological Service description and applications
Third European Conference on Radar Meteorology (ERAD), 2004Co-Authors: Joan Bech, E Vilaclara, Nicolau Pineda, Tomeu Rigo, J M Lopez, F Ohora, J Lorente, D Sempere, F X FabregasAbstract:In autumn 2003 a new radar system was installed and included in the radar network of the Catalan Meteorological Service (SMC). It was the third unit of a network of four radars designed to cover the complex topographical area of Catalonia (approximately 32 000 square km), located in the NE of Spain. The initial design of the network was performed in 1997 considering the already existing weather radar installed in Vallirana (near Barcelona) in 1996. Using a propagation model, visibility maps were derived to study new potential radar sites aiming to minimize beam blockage problems to obtain an improved coverage area suitable for radar quantitative precipitation estimates. The high density of the radar network was calculated to improve the precipitation observing capacity in a highly complex topographical area prone to flooding according to the dominant torrential Mediterranean regime. From the initial network proposal, three new sites were evaluated and selected in the centre of Catalonia and in the northern and southern coast. Two new radars were installed in 2002 and 2003 and the remaining one is planned to be built during 2005. Some hardware characteristics of the first two systems were upgraded during 2003 so the radars are very similar from the technical point of view. They have an offset 4 m antenna dish providing a one-degree main lobe beamwidth. The antenna controller is the RCP-8 unit manufactured by Sigmet, Inc. The C-band transmitter, is based in a Travelling Wave Tube design, and is controlled by a digital Sigmet RVP-8 processor and receiver. Pulse compression is being implemented in 2004 to allow the use of long pulses with radial range resolutions similar to those obtained with higher power transmitters. The main application of the radar network is weather surveillance and monitoring of heavy precipitation events by the Catalan Meteorological Service in combination with other observational tools (lightning detection system, satellite images, etc.). Other applications, developed through collaboration research projects, include the assimilation of radar observations to improve preCorrespondence to: J. Bech (jbech@meteocat.com) cipitation forecasts of mesoscale NWP models and also the enhanced processing of radar data to allow its quantitative use in a specifically-designed hydrological model.
Gonca Ozmen Koca - One of the best experts on this subject based on the ideXlab platform.
-
estimation of solar radiation using artificial neural networks with different input parameters for mediterranean region of anatolia in turkey
Expert Systems With Applications, 2011Co-Authors: Ahmet Koca, Hakan F Oztop, Yasin Varol, Gonca Ozmen KocaAbstract:An artificial neural network (ANN) model was used to estimate the solar radiation parameters for seven cities from Mediterranean region of Anatolia in Turkey. As well known that Turkey is a bridge between Asia and Europe and it lies in a sunny belt, between 36^o and 42^oN latitudes. Indeed, the country has sufficient solar radiation intensities for solar applications. In order to make estimation of solar radiation, the data from the Turkish State and Meteorological Service were used. Data of 2006 were used for testing and data of 2005, 2007, and 2008 were estimated. Effects of number of input parameters were tested on solar radiation that was output layer. With this aim, number of input layer parameters changed from 2 to 6. The obtained results indicated that the method could be used by researchers or scientists to design high efficiency solar devices. It was also found that number of input parameters was the most effective parameter on estimation of future data on solar radiation.