Abstract 021

Abstract ID: 021

Impact of the Initial Stratospheric Polar Vortex State on East Asian Spring Rainfall Prediction in Seasonal Forecast Models

Lead Author: Bowen Liu
Nanjing University of Information Science and Technology, China

Keywords: Stratospheric polar vortex, Spring rainfall, Seasonal forecast models, East Asia

Abstract: The spring seasonal rainfall in East Asia is controlled by both tropical and extratropical climate systems. On the seasonal timescale, the possible downward impact of the initial stratospheric polar vortex state on the East Asian spring rainfall forecasts is still not widely explored in the literature yet, although the role of ENSO has attracted wide attention. Using outputs from seven seasonal forecast models, this study tests the role of the initial stratospheric polar vortex state in forecasting spring rainfall across East Asia. The main conclusions are as follows.
i. Systematic biases are found for the long-term mean stratospheric polar vortex in most seasonal forecast systems, including an oversized area, an overestimated circular shape, and a too-strong strength. These biases are also common in three generations of CMIP models ranging from CMIP3 to CMIP5/6 (Rao et al. 2022). The polar vortex is also centered too close to the North Pole in most models. The climatological longitudinal position of the stratospheric polar vortex centroid also varies with the seasonal forecast model. Out of those metrics, the stratospheric polar vortex strength shows a significant impact on East Asian spring rainfall in observations.
ii. Most models reasonably forecast the interannual variations of the stratospheric polar vortex strength in March with the ACC between forecasts and the ERA5 reanalysis exceeding 0.9 for most models. In contrast, seasonal forecast models show a wide spread in the East Asian spring rainfall forecasts for four representative regions (SE CN, CN-RU, KR, and JP). Few models can possess a reasonable skill of forecasting seasonal rainfall in all regions, although the BCC model shows a relatively high skill for spring rainfall forecasts in most regions (SE CN, CN-RU, and KR).
iii. Most models do not reproduce the linkage between the stratospheric polar vortex in March and East Asian spring rainfall. Specifically, while all models forecast the observed positive NAM response over the pole in spring following an initial strong stratospheric polar vortex, they struggle to represent the anomalous circulation centers in midlatitudes. The tropospheric positive height anomalies over Lake Baikal and its surrounding regions extend farther equatorward in East Asia, corresponding to an anomalous anticyclone and a decrease in spring rainfall over SE CN. Only two of the seven models can produce the dry anomalies in SE CN associated with the strong stratospheric polar vortex in March.
iv. In contrast, the observed wet anomalies in CN-RU, KR, and JP following a strong stratospheric polar vortex are missing in models. The observed wind shear between the anomalous anticyclone over Lake Baikal and another one over North Pacific at 200 hPa is not forecasted in models. Meanwhile, the observed anomalous cyclone in Northeast Asia at 850 hPa in spring is forecasted farther eastward, which explains the low skill of producing the wet anomalies in CN-RU, KR, and JP by models.
v. The year-by-year ENSO SST in March is comparably forecasted by models with the ACC between forecasts and ERSSTv5 exceeding 0.9 for all models. All models can forecast the wet pattern in East Asia during the warm ENSO spring. The observed anomalous anticyclone over the Northwestern Pacific and the southwesterly anomalies along the coastal China, Korea, and Japan are also forecasted by models. In contrast, the anomalous cyclone over Northeast Asia and the wet anomalies in CN-RU is missing in most models.
vi. Real-time forecasts in spring of 2020 and 2021 are selected to validate the inter-model relationship between the forecasted polar vortex bias and East Asian rainfall bias. The negative relationship between the vortex strength and spring rainfall is also identified in parts of East Asia. Similarly, the positive relationship between the ENSO SST and the spring rainfall in parts of East Asia is further verified.

Jian Rao, Tongwen Wu, Chaim I. Garfinkel, Jingjia Luo, Yixiong Lu, Min Chu, Jinggao Hu