Improving Trigger Functions for Convective Parameterization Schemes Using GOAmazon Observations

Using observations from the Green Ocean Amazon (GOAmazon) field campaign, this study aims to improve trigger functions of convection schemes. Results show that the CAPE generation rate (dCAPE)-type triggers are the first tier and that the Bechtold and heated condensation framework (HCF) triggers are a distant second tier. The composite analysis reveals that the undilute dCAPE trigger underpredicts convection when there is bottom-heavy upward motion but overpredicts convection with low-level downward and upperlevel upward motions. The empirical orthogonal function (EOF) analysis on vertical velocity shows that EOF1 (62.65%) exhibits upward motion throughout the troposphere and that EOF2 (28.05%) has lower-level upward motion and upper-level downward motion. Both of them have close relationships with precipitation, indicating the role of vertical velocity in triggering convection. The skill sensitivity analysis shows that the inclusion of 700-hPa upward motion significantly enhances the undilute dCAPE trigger. For the dilute dCAPE trigger, entrainment rate and dCAPE threshold are optimized to improve it. Opposite to dCAPEtype triggers, the Bechtold trigger overemphasizes the low-level vertical velocity and underpredicts the mature and decaying phases of long-lasting convection events. The HCF trigger overemphasizes the nearsurface moist static energy and overlooks the vertical velocity. The performance of dCAPE-type triggers on various convective systems over the Amazon region is examined. The eastward-propagating systems are best represented, with only a few underpredictions in their decaying stages. The weak locally occurring systems and marginal phases of westward-propagating systems are easy to underpredict. The revised dCAPE-type triggers perform better on different convection systems and the diurnal cycle of convection.