[A-AS02] 気象の予測可能性から制御可能性へ

[AAS02-07] Uncertainty quantification of sub-grid scale parameterizations in atmospheric models toward robust theory of weather controllability

*澤田 洋平¹、Le Duc¹、冨澤 風翔¹、Mai Hung²、廣瀬 郁希¹、斉藤 和雄³ (1.東京大学 工学系研究科、2.ベトナム水文気象庁、3.東京大学 大気海洋研究所)

キーワード：不確実性定量化

Simulated strong sensitivity to initial conditions in atmosphere is a promising clue to realize weather control. However, this simulation of sensitivity to initial conditions is greatly affected by parametric uncertainties in sub-grid scale parameterization schemes in atmospheric models. Most of the current operational ensemble prediction systems perform the integration of an atmospheric model with fixed model parameters from different initial conditions and do not explicitly consider the uncertainties of model parameters, indicating that uncertainty quantification for sub-grid scale parameterizations is still in its infancy. Here we present the current advance of our uncertainty quantification of model parameters in a mesoscale atmospheric model. We realize to estimate non-parametric posterior distribution of parameters in sub-grid scale schemes using observation data with a reasonable computational cost. Three case studies will be provided: (1) some applications to an operational system in Vietnam, (2) severe stationary line shaped precipitation system, and (3) rapid intensification of a tropical cyclone.