Convection‐permitting simulations are used to investigate the key mechanism of Meiyu precipitation diurnal cycle over China. Six days from the 2014 Meiyu season are used to produce a “north” composite rainband over the Yangtze‐Huaihe River Basin and another 6 days used to produce a “south” composite band. Both rainbands have peak rainfall in the early morning, while the south band has a secondary peak in the afternoon. Low‐level ageostrophic winds (AGWs) are found to exhibit diurnal cycles with clockwise rotations and their directions, and magnitudes depend on the background geostrophic monsoon flows. Net moisture flux into a control volume enclosing each rainband is almost purely due to AGWs. For both rainbands, net flux reaches maximum at ~04 LST, about 3–4 hr before morning precipitation peak. For the north band, a prominent minimum occurs at ~19 LST, 4 hr before the precipitation minimum. The moisture fluxes through the southern control volume boundary make the largest contributions to the net flux and its diurnal variations. The diurnal variations of the AGWs and their relationship with the background monsoon flows agree very well with the prediction of Blackadar boundary layer inertial oscillation theory, and the convergence forcing by the AGWs resulting from the inertial oscillations plays a paramount role in modulating the diurnal cycles of Meiyu front precipitation, including the creation of early morning peak and evening minimum. Feedback of latent heat release plays only a secondary role. The commonly recognized diurnal monsoon variability can be explained by the Blackadar inertial oscillation theory.

The impacts of ocean feedback on tropical cyclones (TCs) are investigated using a coupled atmosphere–ocean model under idealized TC and cold core eddy (CCE) conditions. Results reveal negative impacts of the ocean coupling on TC development. The cold wake induced by a TC not only weakens the TC intensity but also limits the expansion of the storm circulation. The presence of CCE has boosted the TC-induced sea surface temperature cooling, which conversely inhibits the TC development. The TC appears to be weakened as it encounters the CCE edge. The intensity reduction attains a maximum shortly after the TC passes over the CCE center, and simultaneously the CCE-induced asymmetry of the storm structure is most significant as well. The TC undergoes a period of recovery after departure from the CCE, lasting about 36–48 h. During this time the residual asymmetry caused by the CCE is smoothed gradually by storm axisymmetrization. The CCE has induced smaller TC size throughout the simulation even after the TC intensity has completely recovered, an indication of longer recovery time for the TC size. Notably cooler and moister eye air in the lower troposphere, just under the warm-core height, is found in the experiment with CCE. The water vapor mixing ratio budget analysis indicates that it is primarily attributed to changes in vertical advection that occurred in the eye, that is, the undermined eye subsidence associated with the suppressed eyewall convection. The horizontal patterns of vertical motion in the boundary layer are also distinctly changed by the CCE.

On the basis of global positioning system dropsonde data, Japan Meteorology Agency Regional Spectral Model analysis data, National Centers for Environmental Prediction reanalysis data, satellite products from the Naval Research Laboratory, and best-track tropical-cyclone (TC) datasets from the Shanghai Typhoon Institute, the statistical characteristics of the ducts induced by TCs (TC ducts) over the western North Pacific Ocean were analyzed for the period from September 2003 to September 2006, and two typical strong-duct cases with remarkable differences in formation cause were analyzed and compared. Of the total of 357 dropsondes, there are 212 cases that show ducting conditions, with an occurrence percentage of ~59%. Of the 212 TC-duct events, profiles with multiple ducting layers make up nearly one-half, with the main type of ducts being elevated ducts; in contrast, weak ducts make up over one-half, resulting in a weak median duct strength and thickness. Ducts formed in the transition zone, especially on the left side of TC tracks, tend to be much stronger and thicker than those formed inside TCs. The former are induced by the interaction between TCs and their surrounding systems, such as the inrush of dry and cold air from the north on the left side of TC tracks. The latter are associated with successive subsidence in the gaps between spiral cloud bands. With increasing TC intensities, the associated ducts inside TCs tend to be much stronger and thicker and to appear at higher altitudes.

Based on the Coupled Ocean-Atmospheric Response Experiment (COARE) bulk algorithm and the Naval Postgraduate School (NPS) model, a universal evaporation duct (UED) model that can flexibly accommodate the latest improvements in component (such as stability function, velocity roughness, and scalar roughness) schemes for different stratification and wind conditions, is proposed in this paper. With the UED model, the sensitivity of the model-derived evaporation duct height (EDH) to stability function (ψ), ocean wave effect under moderate to high wind speeds, and scalar roughness length parameterization, is investigated, and relative contributions of these factors are compared. The results show that the stability function is a key factor influencing the simulated EDH values. Under unstable conditions, the EDH values from stability functions of Fairall et al. (1996) and Hu and Zhang (1992) are generally higher than those from others; while under stable conditions, unreasonably high EDHs can be avoided by use of the stability functions of Hu and Zhang (1992) and Grachev et al. (2007). Under moderate to high wind speeds, the increase in velocity roughness length z ₀ due to consideration of the true ocean wave effect acts to reduce modeled EDH values; this trend is more pronounced under stable conditions. Although the scalar roughness length parameterization has a minor effect on the model-derived EDH, a positive correlation is found between the scalar roughness length z _0q and the model-derived EDH.