Climatological Characteristics and Predictability of Tropical Cyclone Size

Abstract

The thesis proposes a methodology to estimate tropical cyclone (TC) size from reanalysis data to establish a 31-year TC size climatology to investigate the decadal, annual and seasonal variations of TC size, and the mechanisms of TC size change.

Part 1 proposes and validates a method to use an azimuthally-averaged wind speed at the outer-core of a TC to estimate its size, defined as the azimuthally average of 17 m s^-1 surface wind (R17), from the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR). The validation compares these estimations with homogeneous TC samples from the Quick Scatterometer (QuikSCAT) over the Western North Pacific (WNP) and North Atlantic (NA) based on both dependent and independent samples. The results show that the sizes estimated from CFSR are highly correlated with those inferred from QuikSCAT (r = 0.861). The size-frequency distributions as well as monthly and annual variations from both datasets agree well with each other. The estimated values are very close to the observed values, the mean absolute error and mean square error are, respectively, 0.409° and 0.314° latitude.

Based on the method in Part I, Part II presents the 31-year TC size climatologies for the WNP and NA. The mean TC sizes over the WNP and NA are 2.01° and 1.61° latitude, respectively. TC size in both basins exhibits temporal intra-seasonal and annual as well as spatial variations. In the WNP, the monthly mean size peaks are found in September and October over the WNP, while the peak in the NA is found in November, both of which pass the 95% confidence interval of the Student’s t distribution. TC mean size in the WNP in the 2000s is generally smaller than that in the 1980s and 1990s, decreasing from 2.02 °latitude from 1980‒1999 to 1.87° latitude during 2000‒2010.

The inter-annual TC size variations in the WNP are highly correlated (r =0.816) with El Niño-Southern Oscillation events. In the WNP, the highest percentage of large TCs occurs near 25°N. Such results agree with previous modelling and observational studies.

In Part III, reanalysis data and operational TC forecast model data are used to investigate the effect of angular momentum import and humidity in control TC size change. Using the reanalysis data, TC size change is found to be related to the angular momentum import in the lower troposphere. which suggests that higher angular momentum import in the lower troposphere tends to produce a larger TC. However, changes in humidity bear no obvious relationship with those of TC size.

TC cases in the operational model, ACCESS-TC, from the Australian Bureau of Meteorology are used to further investigate whether the model can reproduce the observational results. The cases with good TC size prediction show that the TC size change is indeed related to the angular momentum import in the lower troposphere in the outer core region while the intensity is related to the angular momentum import near the TC center in lower troposphere and the angular momentum export in the upper troposphere. The results reaffirm that the angular momentum import is an important factor to control the TC size change. Similar to observations, the model predictions also do not show any clear relationship between humidity and TC size change.

Date of Award	20 Jul 2018
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Chung Leung Johnny CHAN (Supervisor)