Winter precipitation predictability in Central Southwest Asia and its representation in seasonal forecast systems

Abstract In Central Southwest Asia (CSWA), winter precipitation contributes significantly to the annual mean, but substantial interannual variations exist. Dynamical models exhibit subpar predictability in this region, but it still needs to be established what limits their skills. Here, we identify the major tropical and extratropical forcings that explain three-fourths of area-averaged seasonal variability in CSWA winter precipitation. Tropical forcing comes from the indirect ENSO pathway, the leading mode of precipitation variability in the tropical Indian Ocean. This precipitation mode is coupled with ENSOrelated sea surface temperature variability in the Pacific Ocean. A direct ENSO influence on CSWA does not extend beyond its Indian Ocean connection. Extratropical forcing comes from a large-scale mode arising from internal atmospheric variability. The spatial structure, interannual variability of tropical forcing, and its teleconnection with CSWA winter precipitation are skillfully depicted in two seasonal forecasting systems, SEAS5 and SPEAR. Extratropical forcing spatial structure is also produced skillfully in the two modeling systems; however, the representation of its interannual seasonal variability and teleconnection with CSWA winter precipitation require improvements. While SEAS5 displays reasonable skill in representing extratropical forcing influence on CSWA winter precipitation and marginal skill in reproducing interannual seasonal variability, SPEAR has negligible ability in both areas. Consequently, these models have limited predictive skills over CSWA in winter. While improvements in representing extratropical forcing in dynamical models may be inherently limited as it primarily arises from internal