Multivariate Postprocessing Methods for High-Dimensional Seasonal Weather Forecasts

Heinrich, Claudio; Hellton, Kristoffer H.; Lenkoski, Alex; Thorarinsdottir, Thordis L.

doi:10.1080/01621459.2020.1769634

Cited by 12 publications

(14 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…given in Section 4.1 . We apply a permutation test (Heinrich et al, 2020 ) to the scores to test if the mean differences are significantly different. The corresponding scores for MOGREPS-UK wind speed forecasts and the two EMOS variants applied are shown in Table 3.2 .…”

Section: Raft For Ensemble Mean Forecastsmentioning

confidence: 99%

“…It is also possible to use less empirical and more formal tests like the Diebold-Mariano test . Heinrich et al ( 2020 ) propose a permutation test that has the advantage that the asymptotic variance does not have to be estimated. We use this permutation test in Paper II to compare the performance of different combinations of post-processing methods.…”

Section: Forecast Verificationmentioning

confidence: 99%

“…In this table only scores for RAFT forecasts that have been adjusted 1 h previously are shown, i.e., the optimal forecasts. We test the significance of score differences by applying a permutation test based on resampling, as described in Heinrich et al (2019) and . Both EMOS methods increase the deterministic skill considerably when compared to the raw ensemble and then are further improved by applying RAFT m .…”

Section: Choice Of Emos Modelmentioning

confidence: 99%

“…Van Schaeybroeck and Vannitsem, 2018), or other slowly-evolving components of the Earth system such as sea-surface temperature (e.g. Heinrich et al, 2019). However, forecast users commonly need information on atmospheric variables such as surface temperature and precipitation (Roulin and Vannitsem, 2019).…”

Section: Title Trajectory Adjustment Of Lagged Seasonal Forecast Ensembles 1 Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Verification: Assessment of Calibration and Accuracy

Thorarinsdottir

Schuhen

2018

Statistical Postprocessing of Ensemble Forecasts

Self Cite

View full text Add to dashboard Cite

Rapid adjustment and post-processing of temperature forecast trajectories II Order of operation for multi-stage post-processing of ensemble wind forecast trajectories III Trajectory adjustment of lagged seasonal forecast ensembles 121 IV Verification: assessment of calibration and accuracy 137 vi List of Tables 3.1 Continuous ranked probability score and root-mean-square error (in°C ) of raw and post-processed MOGREPS-UK temperature forecasts, averaged over all forecast cases from the 15 UTC run in the test set. All pairwise score differences are statistically significant at α = 0.01. . . . . 20 3.2 Continuous ranked probability score and root-mean-square error (in knots) of raw and post-processed MOGREPS-UK wind speed forecasts, averaged over all forecast cases from the 15 UTC run in the test set. All pairwise score differences are statistically significant at α = 0.01. . . . . 20 3.3 Continuous ranked probability score and root-mean-square error (in knots) of different combinations of EMOS and RAFT methods, averaged over all forecast cases from the 15 UTC run in the test set. The RAFT forecasts are taken from the final adjustments for each lead time. All pairwise score differences are statistically significant at α = 0.05, apart from the RMSE differences between identical combinations using gEMOS

show abstract

Section: Raft For Ensemble Mean Forecastsmentioning

confidence: 99%

Section: Forecast Verificationmentioning

confidence: 99%

Section: Choice Of Emos Modelmentioning

confidence: 99%

Section: Title Trajectory Adjustment Of Lagged Seasonal Forecast Ensembles 1 Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Verification: Assessment of Calibration and Accuracy

Thorarinsdottir

Schuhen

2018

Statistical Postprocessing of Ensemble Forecasts

Self Cite

View full text Add to dashboard Cite

show abstract

“…Model evaluation is performed using the probability integral transform (PIT, see e.g. Heinrich et al, 2020) and the continuous ranked probability score (CRPS, Matheson & Winkler, 1976). A closed-form expression for the CRPS with an FPLD forecast is also developed.…”

Section: Introductionmentioning

confidence: 99%

Quantile based modelling of diurnal temperature range with the five-parameter lambda distribution

Vandeskog,

Thorarinsdottir,

Steinsland

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Diurnal temperature range is an important variable in climate science that can provide information regarding climate variability and climate change. Changes in diurnal temperature range can have implications for human health, ecology and hydrology, among others. Yet, the statistical literature on modelling diurnal temperature range is lacking. This paper proposes to model the distribution of diurnal temperature range using the five-parameter lambda distribution (FPLD). Additionally, in order to model diurnal temperature range with explanatory variables, we propose a distributional quantile regression model that combines quantile regression with marginal modelling using the FPLD. Inference is performed using the method of quantiles. The models are fitted to 30 years of daily observations of diurnal temperature range from 112 weather stations in the southern part of Norway. The flexible FPLD shows great promise as a model for diurnal temperature range, and performs well against competing models. The regression model is fitted to diurnal temperature range data using geographic, orographic and climatological explanatory variables. It performs well and captures much of the spatial variation in the distribution of diurnal temperature range in Norway.

show abstract

Quantile based modeling of diurnal temperature range with the five‐parameter lambda distribution

et al. 2022

Self Cite

View full text Add to dashboard Cite

Diurnal temperature range is an important variable in climate science that can provide information regarding climate variability and climate change. Changes in diurnal temperature range can have implications for hydrology, human health and ecology, among others. Yet, the statistical literature on modeling diurnal temperature range is lacking. In this article we propose to model the distribution of diurnal temperature range using the five-parameter lambda (FPL) distribution. Additionally, in order to model diurnal temperature range with explanatory variables, we propose a distributional quantile regression model that combines quantile regression with marginal modeling using the FPL distribution. Inference is performed using the method of quantiles. The models are fitted to 30 years of daily observations of diurnal temperature range from 112 weather stations in the southern part of Norway. The flexible FPL distribution shows great promise as a model for diurnal temperature range, and performs well against competing models. The distributional quantile regression model is fitted to diurnal temperature range data using geographic, orographic, and climatological explanatory variables. It performs well and captures much of the spatial variation in the distribution of diurnal temperature range in Norway.

show abstract

Multivariate Postprocessing Methods for High-Dimensional Seasonal Weather Forecasts

Cited by 12 publications

References 38 publications

Verification: Assessment of Calibration and Accuracy

Verification: Assessment of Calibration and Accuracy

Quantile based modelling of diurnal temperature range with the five-parameter lambda distribution

Quantile based modeling of diurnal temperature range with the five‐parameter lambda distribution

Contact Info

Product

Resources

About