Skill decreases in real-time seasonal climate prediction due to decadal variability

Skillful seasonal prediction of the 2022–23 mega soil drought over the Yangtze River basin by combining dynamical climate prediction and copula analysis

“…However, we also found that the NMME models have large uncertainties in the precipitation prediction as the lead time increases due to the inherently chaotic nature of the atmosphere (Lavers et al 2009, Yuan et al 2011, Smith et al 2012, and the prediction of drought onset quickly loses skill after one month. Given that the sea-air teleconnection and land-atmosphere coupling are key processes related to the generation of extreme events (Taylor et al 2012, Hao et al 2018, Shao et al 2023, more accurate simulations of the atmospheric circulation response to sea surface temperature Hoerling 2014, Schubert et al 2016), as well as further taking into account land-atmosphere coupling that facilitates the propagation of meteorological drought to soil moisture drought may provide prospects for the prediction of soil drought onset at long lead times (Schubert et al 2007, Hao et al 2018, which will benefit a variety of sectors by allowing sufficient time for drought mitigation efforts. Meanwhile, since the catchment memory, that is, prevailing catchment storage capacity (soil water, groundwater), plays an important role in explaining the performance of drought events (Van Hateren et al 2019, Sutanto andVan Lanen 2022), a reasonable reflection of catchment memory in the dynamic-statistical framework may further improve the predictive skill of soil drought.…”