Importance of Environmental Conditions for the Sensitivity of GPS RO Data Assimilation on Tropical Cyclone Formation Simulation

This study analyzes the impact of Global Positioning System (GPS) Radio Occultation (RO) data assimilation on the simulation of tropical cyclone (TC) formation under different large-scale environments in the western North Pacific. The 3-Dimensional Variational (3DVAR) data assimilation method of the Weather Research and Forecasting Model (WRF) is used to evaluate the influence of GPS RO data on TC formation simulation. The GPS RO soundings are obtained from the Constellation Observing Systems for Meteorology, Ionosphere, and Climate / Formosa Satellite Mission 3 (COSMIC / FORMOSAT-3). In each simulation, the cycling assimilation of nonlocal operator refractivity data is performed over a 3-day period starting from 5 days before TC formation, and then the 2-day simulation is proceeded for TC formation. During July−October 2006−2010, 16 TCs are selected and simulated in this study. Of these, 6 TCs develop and form in monsoon environments, 8 in easterly environments, and 2 across easterly and monsoon environments. The formations of 6 TCs can be captured in both simulations with and without assimilation of RO data (Type 1), 6 TCs cannot be captured in the simulations either with or without assimilation of RO data (Type 2), and 4 TCs can be captured only in the simulations with assimilation of RO data (Type 3). The simulation of Type-3 TC formation is sensitive to RO data assimilation. Comparing the environmental conditions among different types of TC shows that Type-1 TCs mainly develop in monsoon environments and across multi-environment, which have higher relative humidity at mid-levels and higher maximum potential intensity before TC formation than other TCs. Type-2 TCs mainly develop in easterly environments, which have the worst environmental conditions (e.g., the lowest genesis potential index) for TC formation than other TCs. On average, TCs formed in easterly environments have lower mid-level relative humidity but higher mid-level vorticity than the total TC average. However, although Type-3 TCs also mainly develop in easterly environments, they have higher relative humidity at mid and high levels before TC formation than other TCs. In other words, the assimilation of RO data can improve the pattern of relative humidity in the few days lead time simulation for Type-3 TCs, which have higher relative humidity originally but is obviously underestimated before TC formation in the simulation without RO data assimilation. This causes the simulation of Type-3 TC formation is sensitive to RO data. More cases, ensemble tests, and COSMIC-II data assimilation will be performed in the future.