28 June 2023 in Earth System Science Data
Statistical postprocessing of medium-range weather forecasts is an important component of modern forecasting systems. Since the beginning of modern data science, numerous new postprocessing methods have been proposed, complementing an already very diverse field. However, one of the questions that frequently arises when considering different methods in the framework of implementing operational postprocessing is the relative performance of the methods for a given specific task. It is particularly challenging to find or construct a common comprehensive dataset that can be used to perform such comparisons. Here, we introduce the first version of EUPPBench (EUMETNET postprocessing benchmark), a dataset of time-aligned forecasts and observations, with the aim to facilitate and standardize this process. This dataset is publicly available at https://github.com/EUPP-benchmark/climetlab-eumetnet-postprocessing-benchmark (31 December 2022) and on Zenodo (https://doi.org/10.5281/zenodo.7429236, Demaeyer, 2022b and https://doi.org/10.5281/zenodo.7708362, Bhend et al., 2023). We provide examples showing how to download and use the data, we propose a set of evaluation methods, and we perform a first benchmark of several methods for the correction of 2 m temperature forecasts.
1 July 2023 in Quarterly Journal of the Royal Meteorological Society
Current practice in predicting future weather is the use of numerical weather prediction (NWP) models to produce ensemble forecasts. Despite of enormous improvements over the last few decades, they still tend to exhibit bias and dispersion errors and, consequently, lack calibration. Therefore, these forecasts may be improved by statistical postprocessing. In this work, we propose a D‐vine‐copula‐based postprocessing for 10 m surface wind speed ensemble forecasts. This approach makes use of quantile regression related to D‐vine copulas, which is highly data driven and allows one to adopt more general dependence structures as the state‐of‐the‐art zero‐truncated ensemble model output statistic (tEMOS) model. We compare local and global D‐vine copula quantile regression (DVQR) models to the corresponding tEMOS models and their gradient‐boosting extensions (tEMOS‐GB) for different sets of predictor variables using one lead time and 60 surface weather stations in Germany. Furthermore, we investigate which types of training periods can improve the performance of tEMOS and the D‐vine‐copula‐based method for wind speed postprocessing. We observe that the D‐vine‐based postprocessing yields a comparable performance with respect to tEMOS if only wind speed ensemble variables are included and to substantial refinements if additional meteorological and station‐specific weather variables are integrated. As our main result, we note that, in the global setting, DVQR is able to provide better scores than tEMOS‐GB in general, whereas the local DVQR is able to substantially outperform the local tEMOS‐GB at particular stations admitting nonlinear relationships among the variables. In addition, we remark that training periods capturing seasonal patterns perform the best. Last but not least, we adapt a criterion for calculating the variable importance in D‐vine copulas and we outline which predictor variables are due to this approach the most important for the correction of wind speed ensemble forecasts.
5 December 2024 in Environmetrics
Vine copulas are flexible dependence models using bivariate copulas as building blocks. If the parameters of the bivariate copulas in the vine copula depend on covariates, one obtains a conditional vine copula. We propose an extension for the estimation of continuous conditional vine copulas, where the parameters of continuous conditional bivariate copulas are estimated sequentially and separately via gradient‐boosting. For this purpose, we link covariates via generalized linear models (GLMs) to Kendall's correlation coefficient from which the corresponding copula parameter can be obtained. In a second step, an additional covariate deselection procedure is applied. The performance of the gradient‐boosted conditional vine copulas is illustrated in a simulation study. Linear covariate effects in low‐ and high‐dimensional settings are investigated separately for the conditional bivariate copulas and the conditional vine copulas. Moreover, the gradient‐boosted conditional vine copulas are applied to the multivariate postprocessing of ensemble weather forecasts in a low‐dimensional covariate setting. The results show that our suggested method is able to outperform the benchmark methods and identifies temporal correlations better. Additionally, we provide an R‐package called boostCopula for this method.
9 September 2024 in Quarterly Journal of the Royal Meteorological Society
The uncertainty in numerical weather prediction models is nowadays quantified by the use of ensemble forecasts. Although these forecasts are continuously improved, they still suffer from systematic bias and dispersion errors. Statistical postprocessing methods, such as the ensemble model output statistics (EMOS), have been shown to substantially correct the forecasts. This work proposes an extension of EMOS in a time‐series framework. Besides taking account of seasonality and trend in the location and scale parameter of the predictive distribution, the autoregressive process in the mean forecast errors or the standardized forecast errors is considered. The models can be further extended by allowing generalized autoregressive conditional heteroscedasticity. Furthermore, it is outlined how to use these models for arbitrary forecast horizons. To illustrate the performance of the suggested EMOS models in time‐series fashion, we present a case study for the postprocessing of 2 m surface temperature forecasts using five different lead times and a set of observation stations in Germany. The results indicate that the time‐series EMOS extensions are able to significantly outperform the benchmark models EMOS and autoregressive EMOS (AR‐EMOS) in most of the lead time–station cases.
21 February 2025 in Journal of the Royal Statistical Society Series C: Applied Statistics
The D-vine copula-based quantile regression (DVQR) is a powerful tool for weather forecasting, as it is able to select informative predictor variables from a large set and takes account of nonlinear relationships among them. However, DVQR shows in its current form a lack in adaptively modelling strongly varying effects among variables, such as temporal and/or spatial effects. Consequently, we propose an extension of the current DVQR, where we specify the parameters of the bivariate copulas in the D-vine copula through Kendall’s τ to which additional covariates are linked. The parameterization of the correlation parameter allows generalized additive models (GAMs) to incorporate, e.g. linear, nonlinear, and spatial effects as well as interactions. The new method is called GAM-DVQR, and its performance is illustrated in a case study on postprocessing 2 m surface temperature ensemble weather forecasts. We investigate constant as well as time-dependent Kendall’s τ correlation models. The results indicate that the GAM-DVQR models are able to identify time-dependent correlations and significantly outperform state-of-the-art postprocessing methods. Furthermore, the introduced temporal parameterization allows a more economical and faster model estimation in comparison to DVQR using a sliding training window. To complement this article, we provide an R-package for our method called gamvinereg.