Contributions of Surface Sensible Heat Fluxes to Tropical Cyclone Intensity

1. Introduction Tropical cyclones (TCs) are one of the most hazardous tropical weather systems (Simpson et al. 2002). After originated over the warm ocean surface, tropical cyclonic disturbances could grow to TCs by inducing upward transport of heat fluxes that are positively related to surface wind speeds (Emanuel 1986). The supply of surface heat fluxes, with both flux forms of sensible heat (SHX) and latent heat (LHX), is well identified to be the energy source for the intensification of TCs (Malkus and Riehl 1960). Given that the LHX and SHX are often added up together to explain the changes in TC intensity (e.g., Chan et al. 2001; Chen et al. 2010), the specific roles of SHX in TC evolution are still not well understood. This study aims to explore the contributions of SHX to TC by conducting idealized numerical simulations. 2. Model setup and experimental design This study uses the Weather Research and Forecasting (WRF) model. The model is set on an f plane at 20°N, containing three domains with horizontal resolutions of 15 km, 5 km and 1.67 km, and dimensions of 220×220, 202×202 and 301×301, respectively. The quiescent and horizontally uniform environmental fields are specified at all levels according to the mean tropical sounding of temperature and humidity profiles given by Jordan (1958). A bogus vortex in hydrostatic and gradient wind balance is implanted in the idealized atmosphere environment with a maximum wind speed of 24 m s-1. The WRF model is first integrated for 24 h as the spinup. SHX and LHX are calculated with the following equations: SHX=RATIO1*FLHC(θg -θa), (3) LHX=RATIO2*FLQC(qg -qa), (4) where FLHC and FLQC are intermediate coefficients in the model; RATIO1 and RATIO2 are 1 in default. Since the surface heat fluxes serve as the bottom boundary conditions for YSU scheme while the surface exchange coefficients are not the input arguments, the effects of SHX can be effectively manipulated by setting RATIO1 to different ratios. Accordingly, several experiments are conducted to investigate the contributions of SHX to TC evolution by changing the values of RATIO1. The experiment with default values of RATIO1 and RATIO2 is considered as the control run, denoted as CTRL. The experiment with the RATIO1 set to 0.5, namely with a half of SHX removed, is denoted as HALF. The experiment with the RATIO1 set to zero is denoted as NOSH, in which the SHX are completely excluded. An experiment with LHX completely removed (by setting RATIO2 to zero) is also conducted as a comparison, denoted as NOLH. All runs are integrated for a total of 96 h. 3. Results Since originated in quiescent environment on an f plane, the storm keeps basically stationary throughout the simulation (not shown). Figure 1 shows the evolution of storm intensity and size in terms of the minimum sea level pressure, the maximum azimuthal mean wind speed at the lowest model level (about 30 m in altitude), and the radius of hurricane-force (33 m s-1) wind. During the spinup period the storm intensifies rapidly and expands steadily, with the moist processes, low-level inflow, and upper-level outflow established progressively (not shown). Thereafter the storms in all runs but NOLH intensify relatively slowly, while their circulations still enlarge with time. At about 84 h the storm in CTRL basically reaches its mature state, with a minimum sea level pressure of 924 hPa and a maximum surface wind of 55 m s-1. Of interest is that although SHX are one factor of the energy sources for TCs, the influences of cutting down a half of SHX or completely removing the SHX are nonetheless not significant on storm intensity. Compared with CTRL, both HALF and NOSH give similar storm intensities with small fluctuations present. In the first 24 h, the storm in NOSH is slightly weaker than that in CTRL, with the storm central pressure difference less than 5 hPa, and the surface wind speed difference less than 5 m s-1. From about 24 h to 60 h NOSH even gives a more intensified storm than CTRL. Afterward the storm in CTRL is again stronger than that in NOSH. The storm intensity in HALF shows slightly different evolutionary trend from that in NOSH, but remains quite close to that in CTRL, especially in the first 36 h. Contrasting with the behavior of storm intensity, the expansion of storm size in NOSH is smoother than in CTRL, followed by that in HALF. As a consequence, their storm size differences tend to be enlarged in the later stage. The size difference between NOSH and HALF seems to be much larger than that between HALF and CTRL, in that at the end of the simulation the storm in NOSH is roughly 23 km (21%) shrunk compared with that in CTRL, while the storm size in HALF is reduced by merely 7 km (6.4%) relative to that in CTRL. This implies that the influences of SHX on storm size may be not linearly dependent on their magnitudes. The LHX are therefore indicated to be the dominant energy source for TC intensification, whereas the SHX are not. 4. Summary and discussion The cloud resolving simulations with and without considering the SHX are conducted and compared to investigate the roles of SHX during TC evolution. A predominant finding is that the influences of SHX on TC intensity are not significant. Specifically, after completely removing the SHX, the storm intensity is slightly weakened in the early period of about 24 h, with its central pressure elevated by less than 5 hPa. Afterward the storm intensity evolves with small fluctuations, and could be even stronger than the control run. Contrasting to the behavior of TC intensity, removing the upward transfer of SHX has led to evidently shrinkage of the TC size, that the radius of hurricane-force winds is decreased by roughly 21%. Acknowledgements. This work is supported by the National Natural Science Foundation of China with Grant 41230421, and the 973 project (2015CB452802) of the Ministry of Science and Technology