Probabilistic Precipitation Forecasting over East Asia Using Bayesian Model Averaging

Abstract

Bayesian model averaging (BMA) was applied to improve the prediction skill of 1–15-day, 24-h accumulated precipitation over East Asia based on the ensemble prediction system (EPS) outputs of ECMWF, NCEP, and UKMO from the TIGGE datasets. Standard BMA deterministic forecasts were accurate for light-precipitation events but with limited ability for moderate- and heavy-precipitation events. The categorized BMA model based on precipitation categories was proposed to improve the BMA capacity for moderate and heavy precipitation in this study. Results showed that the categorized BMA deterministic forecasts were superior to the standard one, especially for moderate and heavy precipitation. The categorized BMA also provided a better calibrated probability of precipitation and a sharper prediction probability density function than the standard one and the raw ensembles. Moreover, BMA forecasts based on multimodel EPSs outperformed those based on a single-model EPS for all lead times. Comparisons between the two BMA models, logistic regression, and raw ensemble forecasts for probabilistic precipitation forecasts illustrated that the categorized BMA method performed best. For 10–15-day extended-range probabilistic forecasts, the initial BMA performances were inferior to the climatology forecasts, while they became much better after preprocessing the initial data with the running mean method. With increasing running steps, the BMA model generally had better performance for light to moderate precipitation but had limited ability for heavy precipitation. In general, the categorized BMA methodology combined with the running mean method improved the prediction skill of 1–15-day, 24-h accumulated precipitation over East Asia.

Comparative Study on Probabilistic Forecasts of Heavy Rainfall in Mountainous Areas of the Wujiang River Basin in China Based on TIGGE Data

“…At present, statistical post-processing methods mainly include the probability matched mean [9], the logistic regression method (LR) [10,11], the neighborhood method [12,13], the multi-variate Gaussian ensemble kernel dressing method [14], the object probability prediction method [15], the ensemble pseudo-bias correction [16], the Bayesian model averaging method (BMA) [17][18][19], and the empirical connection method [20]. The current research results show that the logical regression [21][22][23], the nonlinear Gaussian fitting method, and the BMA [24][25][26][27][28][29][30][31][32][33][34] can provide comparatively ideal forecast results, and are the main ones studied.…”

Calibration Of Rainfall Ensemble Prediction Of ECMWF System 4 Using Bayesian Model Averaging

“…Selain melakukan koreksi bias dengan gQM, tahapan utama pada penelitian ini adalah menerapkan BMA pada RAW model ECS4. BMA adalah metode post processing statistik yang menghasilkan prediksi probabilistik dari prediksi ensemble dalam bentuk fungsi kepadatan peluang atau Probability Density Function (PDF) prediktif (Abraham and Puthiyidam, 2016; Baran et al, 2019;Ji et al, 2019;Liu et al, 2019;Song et al, 2018;Xu et al, 2019 Normal (Raftery et al, 2005), curah hujan harian dengan gamma nol (Sloughter et al, 2007), dan kecepatan angin dengan gamma (Sloughter et al, 2010). Menurut Gneiting (2014) dan Sloughter et al (2007), distribusi PDF prediktif BMA untuk curah hujan harian adalah gamma nol.…”

Probabilistic Forecasting of the 500 hPa Geopotential Height over the Northern Hemisphere Using TIGGE Multi-model Ensemble Forecasts

“…The predictive parameters in EMOS are estimated using multiple regression equations with the raw ensemble forecasts and the corresponding verifications over a training period. Both BMA and EMOS have been applied to several different weather quantities (i.e., temperature, wind, and precipitation) and have been shown to improve prediction skills relative to the raw ensembles [34][35][36][37][38][39][40][41][42][43][44][45][46][47].…”