Probabilistic forecasting based on ensemble forecasts and EMOS method for TGR inflow

Abstract

Probabilistic inflow forecasts can quantify the uncertainty involved in the forecasting process and provide useful risk information for reservoir management. This study proposed a probabilistic inflow forecasting scheme for the Three Gorges Reservoir (TGR) at 1–3 d lead times. The post-processing method Ensemble Model Output Statistics (EMOS) is used to derive probabilistic inflow forecasts from ensemble inflow forecasts. Considering the inherent skew feature of the inflow series, lognormal and gamma distributions are used as EMOS predictive distributions in addition to conventional normal distribution. Results show that TGR’s ensemble inflow forecasts at 1–3 d lead times perform well with high model efficiency and small mean absolute error. Underestimation of forecasting uncertainty is observed for the raw ensemble inflow forecasts with biased probability integral transform (PIT) histograms. The three EMOS probabilistic forecasts outperform the raw ensemble forecasts in terms of both deterministic and probabilistic performance at 1–3 d lead times. The EMOS results are more reliable with much flatter PIT histograms, coverage rates approximate to the nominal coverage 89.47% and satisfactory sharpness. Results also show that EMOS with gamma distribution is superior to normal and lognormal distributions. This research can provide reliable probabilistic inflow forecasts without much variation of TGR’s operational inflow forecasting procedure.

Keywords

ensemble forecast / probabilistic forecast / numeric weather prediction / EMOS / Three Gorges Reservoir

References

[1] Arnal L, Ramos M H, de Perez E C, Cloke H L, Stephens E, Wetterhall F, van Andel S J, Pappenberger F (2016). Willingness-to-pay for a probabilistic flood forecast: a risk-based decision-making game. Hydrol Earth Syst Sci, 20(8): 3109–3128

[2] Baran S, Lerch S (2015). Lognormal distribution based EMOS models for probabilistic wind speed forecasting. Q J R Meteorol Soc, 141(691) 2289–2299

[3] Baran S, Nemoda D (2016). Censored and shifted gamma distribution based EMOS model for probabilistic quantitative precipitation forecasting. Environmetrics, 27(5): 280–292

[4] Bourdin D R, Nipen T N, Stull R B (2014). Reliable probabilistic forecasts from an ensemble reservoir inflow forecasting system. Water Resour Res, 50(4): 3108–3130

[5] Bröcker J, Smith L A (2007). Increasing the reliability of reliability diagrams. Weather Forecast, 22(3): 651–661

[6] Chen L, Singh V P, Guo S, Zhou J, Zhang J (2015). Copula-based method for multisite monthly and daily streamflow simulation. J Hydrol (Amst), 528: 369–384

[7] Cloke H L, Pappenberger F (2009). Ensemble flood forecasting: a review. J Hydrol (Amst), 375(3–4): 613–626

[8] Cunge J A (1969). On the subject of a flood propagation computation method (Muskingum method). J Hydraul Res, 7(2): 205–230

[9] Duan Q, Ajami N K, Gao X, Sorooshian S (2007). Multi-model ensemble hydrologic prediction using Bayesian model averaging. Adv Water Resour, 30(5): 1371–1386

[10] Dunne T (1978). Field studies of hillslope flow processes. Hillslope hydrology, 227: 227–293

[11] Emam A R, Kappas M, Fassnacht S, Linh N H K (2018). Uncertainty analysis of hydrological modeling in a tropical area using different algorithms. Front Earth Sci, 12(4): 661–671

[12] Fernandez B, Salas J D (1986). Periodic gamma autoregressive processes for operational hydrology. Water Resour Res, 22(10): 1385–1396

[13] Gneiting T, Balabdaoui F, Raftery A E (2007). Probabilistic forecasts, calibration and sharpness. J R Stat Soc, 69(2): 243–268

[14] Gneiting T, Raftery A E (2005). Weather forecasting with ensemble methods. Science, 310(5746): 248–249

[15] Gneiting T, Katzfuss M (2014). Probabilistic forecasting. J R Stat Soc, 1(1): 125–151

[16] Goldberg D E (1989). Genetic algorithm in search, optimization, and machine learning. Addison Wesley: 2104–2116

[17] Hamill T M (2001). Interpretation of rank histograms for verifying ensemble forecasts. Mon Weather Rev, 129(3): 550–560

[18] Hardy J, Gourley J J, Kirstetter P E, Hong Y, Kong F, Flamig Z L (2016). A method for probabilistic flash flood forecasting. J Hydrol (Amst), 541: 480–494

[19] Hemri S, Lisniak D, Klein B (2015). Multivariate postprocessing techniques for probabilistic hydrological forecasting. Water Resour Res, 51(9): 7436–7451

[20] Huang K D, Ye L, Chen L, Wang Q, Dai L, Zhou J, Singh V P, Huang M, Zhang J (2018). Risk analysis of flood control reservoir operation considering multiple uncertainties. J Hydrol (Amst), 565: 672–684

[21] Hu C H, Guo S L, Xiong L H, Peng D (2005). A modified Xinanjiang model and its application in Northern China. Hydrol Res, 36(2): 175–192

[22] Jiang S, Ren L, Hong Y, Yang X, Ma M, Zhang Y, Yuan F (2014). Improvement of multi-satellite real-time precipitation products for ensemble streamflow simulation in a middle latitude basin in south China. Water Resour Manage, 28(8): 2259–2278

[23] Jie M X, Chen H, Xu C Y, Zeng Q, Chen J, Kim J S, Guo S, Guo F Q (2018). Transferability of conceptual hydrological models across temporal resolutions: approach and application. Water Resour Manage, 32(4): 1367–1381

[24] Kang L, Zhou L, Zhang S (2017). Parameter estimation of two improved nonlinear Muskingum models considering the lateral flow using a hybrid Algorithm. Water Resour Manage, 31(14): 4449–4467

[25] Khan M M, Shamseldin A Y, Melville B W, Shoaib M (2015). Stratification of NWP forecasts for medium-range ensemble streamflow forecasting. J Hydrol Eng, 20(7): 04014076

[26] Laiolo P, Gabellani S, Rebora N, Rudari R, Ferraris L, Ratto S, Stevenin H, Cauduro M (2014). Validation of the flood-proofs probabilistic forecasting system. Hydrol Processes, 28(9): 3466–3481

[27] Lerch S, Thorarinsdottir T L (2013). Comparison of non-homogeneous regression models for probabilistic wind speed forecasting. Tellus, 65(10): 98–110

[28] Lewis D, Singer M J, Dahlgren R A, Tate K W (2000). Hydrology in a California oak woodland watershed: a 17-year study. J Hydrol (Amst), 240(1–2): 106–117

[29] Lan T, Lin K, Liu Z, He Y H, Xu C Y, Zhang H B, Chen X H (2018). A clustering preprocessing framework for the subannual calibration of a hydrological model considering climate-land surface variations. Water Resour Res, 54

[30] Lin K, Lv F, Chen L, Singh V P, Zhang Q, Chen X (2014). Xinanjiang model combined with Curve Number to simulate the effect of land use change on environmental flow. J Hydrol (Amst), 519: 3142–3152

[31] Liu J, Xie Z (2014). BMA probabilistic quantitative precipitation forecasting over the Huaihe Basin using TIGGE multi-model ensemble forecasts. Mon Weather Rev, 142(4): 1542–1555

[32] Liu Z, Guo S, Zhang H, Liu D, Yang G (2016). Comparative study of three updating procedures for real-time flood forecasting. Water Resour Manage, 30(7): 2111–2126

[33] Liu Z, Guo S, Xiong L, Xu C Y (2018). Hydrologic uncertainty processor based on copula function. Hydrol Sci J, 63(1): 74–86

[34] Mascaro G, Vivoni E R, Deidda R (2011). Impact of basin scale and initial condition on ensemble streamflow forecast uncertainty. In: The 25th Conference on Hydrology, American Meteorological Society

[35] Najafi M R, Moradkhani H (2016). Towards ensemble combination of seasonal streamflow forecasts. J Hydrol Eng, 21(1): 04015043

[36] Nash J E, Sutcliffe J V (1970). River flow forecasting through conceptual models: part 1: a discussion of principles. J Hydrol (Amst), 10(3): 282–290

[37] Oudin L, Andréassian V, Mathevet T, Perrin C, Michel C (2006). Dynamic averaging of rainfall-runoff model simulations from complementary model parameterizations. Water Resour Res, 42(7): 887–896

[38] Parasuraman K, Elshorbagy A (2007). Cluster-based hydrologic prediction using Genetic Algorithm-trained Neural Networks. J Hydrol Eng, 12(1): 52–62

[39] Perrin C, Michel C, Andréassian V (2001). Does a large number of parameters enhance model performance? Comparative assessment of common catchment model structures on 429 catchments. J Hydrol (Amst), 242(3–4): 275–301

[40] Perrin C, Michel C, Andréassian V (2003). Improvement of a parsimonious model for streamflow simulation. J Hydrol (Amst), 279(1–4): 275–289

[41] Raftery A E, Gneiting T, Balabdaoui F, Polakowski M (2005). Using Bayesian model averaging to calibrate forecast ensembles. Mon Weather Rev, 133(5): 1155–1174

[42] Steinschneider S, Brown C (2011). Influences of North Atlantic climate variability on low-flows in the Connecticut River Basin. J Hydrol (Amst), 409(1–2): 212–224

[43] Sloughter J M, Raftery A E, Gneiting T, Fraley C (2007). Probabilistic quantitative precipitation forecasting using Bayesian model averaging. Mon Weather Rev, 135(9): 3209–3220

[44] Thorarinsdottir T L, Gneiting T (2010). Probabilistic forecasts of wind speed: ensemble model output statistics by using heteroscedastic censored regression. J R Stat Soc (Ser A), 173(2): 371–388

[45] Tian Y, Xu Y P, Zhang X J (2013). Assessment of climate change impacts on river high flows through comparative use of GR4J, HBV and Xinanjiang models. Water Resour Manage, 27(8): 2871–2888

[46] Todini E (2017). Flood forecasting and decision making in the new millennium: where are we? Water Resour Manage, 31(8): 1–19

[47] Wilks D S, Hamill T M (2007). Comparison of ensemble-MOS methods using GFS reforecasts. Mon Weather Rev, 135(6): 2379–2390

[48] WMO (2005) First Workshop on the THORPEX Interactive Grand Global Ensemble (TIGGE), Final Report

[49] WMO (2010) Workshop on the Strategy and Action Plan of the WMO Flood Forecasting Initiative, Final Report

[50] Wu C L, Chau K W (2006). A flood forecasting neural network model with genetic algorithm. Int J Environ Pollut, 28(3–4): 261–273

[51] Wu Z, Wu J, Lu G (2016). A one-way coupled atmospheric-hydrological modeling system with combination of high-resolution and ensemble precipitation forecasting. Front Earth Sci, 10(3): 432–443

[52] Xiong F, Guo S, Chen L, Yin J, Liu P (2018). Flood frequency analysis using Halphen distribution and maximum entropy. J Hydrol Eng, 23(5): 04018012

[53] Yue S, Ouarda T B M J, Bobée B (2001). A review of bivariate gamma distributions for hydrological application. J Hydrol (Amst), 246(1–4): 1–18

[54] Zhao L, Qi Dan, Tian F, Wu H, D, J, Wang Z, Li A (2012). Probabilistic flood prediction in the upper Huaihe catchment using TIGGE data. J Meteorol Res, 26(1): 62–71

[55] Zhao R (1992). The Xinanjiang model applied in China. J Hydrol (Amst), 135(1–4): 371–381

[56] Zhong Y, Guo S, Ba H, Xiong F, Chang F J, Lin K (2018a). Evaluation of the BMA probabilistic inflow forecasts using TIGGE numeric precipitation predictions based on artificial neural network. Hydrol Res, 49(5): 1417–1433

[57] Zhong Y, Guo S, Liu Z, Wang Y, Yin J (2018b). Quantifying differences between reservoir inflows and dam site floods using frequency and risk analysis methods. Stoch Environ Res Risk Assess, (6):1–15

Acknowledgment

This study is supported by the National Key Research and Development Plan of China (No. 2016YFC0402206) and the National Natural Science Foundation of China (Grant Nos. 51879192, 91647106). Thanks are also given to CWRC for providing necessary data and the three anonymous reviewers’ valuable suggestions to improve our manuscript.

RIGHTS & PERMISSIONS

2019 Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature