On the possible role of condensation-related hydrostatic pressure adjustments in intensification and weakening of tropical cyclones

Abstract

It is shown that condensation and precipitation do not disturb the hydrostatic equilibrium if the local pressure sink (condensation rate expressed in pressure units) is proportional to the local pressure, with a proportionality coefficient $k$ that is independent of altitude. In the real atmosphere, however, the condensation rate depends, among other factors, on the vertical velocity, which varies with height. As a result, condensation generally disturbs hydrostatic equilibrium and induces pressure adjustments through air-mass redistribution. We propose that a profile in which $k$ is maximized in the upper atmosphere leads to additional upward motion and cyclone intensification, whereas a maximum closer to the surface induces downward motion and cyclone weakening. The magnitude of both effects is expected to be set by the strength of the precipitation mass sink. Using observational data, we find that the median intensification and weakening rates -- $12$ and $8$~hPa~day$^{-1}$, respectively, measured over six-hour intervals in Atlantic tropical cyclones -- amount to about three quarters of the maximum concurrent precipitation rate (multiplied by gravity) in the core precipitation region. This implies intensification under conditions of positive vertically integrated air convergence, a regime impossible in modeled dry hurricanes, with the negative pressure tendency arising because precipitation exceeds the vertically integrated moisture convergence by absolute magnitude. The implications of these results for recent studies that evaluate tropical cyclone (de-)intensification using mass continuity equations that neglect the precipitation mass sink are discussed.

Figures (3)

• Figure 1: Intensification rates taken by absolute value $|I|$ and maximum precipitation $gP_m$ in intensifying and weakening storms on land and over the ocean (shown for comparison). Numbers of storms in each group are shown along the lower horizontal axis. Medians of $I$ and $gP_m$ are shown along the upper horizontal axis. Note the logarithmic scale on the vertical axis. Crosses and dots show mean values and outliers, respectively.
• Figure 2: Radius of maximum precipitation $r_P$ and radius of maximum wind $r_m$ in storms where $r_m$ is known. Numbers of storms in each group are the same as in Fig. \ref{['bwc']}. Medians of $r_P$ and $r_m$ are shown along the upper horizontal axis. Note the logarithmic scale on the vertical axis. Crosses and dots show mean values and outliers, respectively.
• Figure 3: Thought experiments illustrating the role of pressure adjustments in shaping precipitation influence on storm intensification rate. Non-condensable and condensable gases are painted pink and blue, respectively. Thin white (black) arrows indicate inflow into, and outflow of gas (condensate) from, the column. Big triangles indicate the direction of pressure adjustment. Yellow spaces indicate pressure perturbations.