Seasonal prediction of high‐resolution temperature at 2‐m height over Mongolia during boreal winter using both coupled general circulation model and artificial neural network

Abstract

The hindcast data of Pusan National University coupled general circulation model (PNU CGCM), a participant model of the Asia‐Pacific Economic Cooperation Climate Center (APCC) Multi‐Model Ensemble Climate Prediction System, and August–October sea‐surface temperature (SST) in the northern Barents–Kara Sea (BKI) and the sea‐ice extent (SIE) in the Chukchi Sea (East Siberian Sea index [ESI]) are used for predicting 20 × 20‐km‐resolution anomalous surface air temperature at 2‐m height (aT2m) over Mongolia for boreal winter. For this purpose, area‐averaged surface air temperature (TI) and sea‐level pressure (SLP) over Mongolia are defined. Then four large‐scale indices, TImdl and SHImdl obtained from PNU CGCM, and TIMLR and SHIMLR obtained from multiple linear regressions on BKI and ESI, are incorporated using the artificial neural network (ANN) method for the prediction and statistical downscaling to obtain the monthly and seasonal 20 × 20‐km‐resolution aT2m over Mongolia in winter. An additional statistical method, which uses BKI and ESI as predictors of TI and SHI together with dynamic prediction by the CGCM, is used because of the relatively low skill of seasonal predictions by most of the state‐of‐the‐art models and the multi‐model ensemble systems over high‐latitude landlocked Eurasian regions such as Mongolia. The results show that the predictabilities of monthly and seasonal 20 × 20‐km‐resolution aT2m over Mongolia in winter are improved by applying ANN to both statistical and dynamical predictions compared to utilizing only dynamic prediction. The predictability gained by the proposed method is also demonstrated by the probabilistic forecast implying that the method forecasts aT2m over Mongolia in winter reasonably well.

Dynamical‐statistical long‐term prediction for tropical cyclone landfalls in East Asia

“…A total of 40 ensemble members is large enough to minimize the initial conditions error and model uncertainty, considering the number of ensemble members produced by global institutes producing long‐term predictions using state‐of‐the‐art CGCMs. The details and performance of the model were introduced and studied by several authors (Ahn and Kim, 2014; Jo and Ahn, 2015; Bayasgalan and Ahn, 2018).…”

“…On both the interannual and interdecadal timescale, the western North Pacific subtropical High (WNPSH) can have a significant impact on the EA TC activities (Choi and Moon, 2012;Wang et al, 2013). Therefore, composite 1991, 1993, 2000, 20181980, 1981, 1983, 1984, 1995, 2003MEA 1994, 2000, 2005, 2006, 20181983, 1986, 1988, 1993, 1995, 1999, 2014SEA 1983, 1994, 19951982, 1999, 2004, 2010, 2014EEA 1991, 1994, 2000, 20181980, 1981, 1983, 1984, 1986, 1998, 2003, 2014 dipole pattern, which is analogous to the positive Pacific-Japan (PJ) teleconnection pattern. In this regard, previous studies showed that TCs land more frequently on NEA when the PJ pattern is positive than negative (Choi et al, 2010a;Kim et al, 2012).…”

Skilful two‐month‐leading hybrid climate prediction for winter temperature over China

“…Prediction models based on these methods have been built for winter temperature in China in operational forecasts (Chen et al, 2007;Ai et al, 2008;Jia et al, 2010;Wang et al, 2013b;Tan et al, 2017), as well as the prediction of extremely cold days Wang, 2016, 2017). Several new statistical approaches have also been developed for temperature forecasts, such as those based on artificial neural networks (Bayasgalan and Ahn, 2018) and spatiotemporal projection models (Vautard et al, 1999;Zhu and Li, 2017). However, the prediction skill of statistical models mainly relies on the statistical stability of the relationships between predictors and predictands, meaning there are considerable uncertainties in the prediction results if statistical relationships mutate (Li et al, 2013).…”