EAPSI: Assessing the impact of atmospheric observations from Taiwanese aircraft reconnaissance on Typhoon forecasts

Finocchio Peter M, Miami FL

Investigators

Abstract

Accurate tropical cyclone forecasts are essential for protecting increasingly vulnerable coastal populations worldwide. However, the accuracy of tropical cyclone forecasts relies on high-quality observations used to initialize weather forecast models, and such observations are particularly sparse over the oceans where these intense cyclones spend most of their lives. To alleviate this under-sampling problem for the numerous typhoons threatening Taiwan each year, Dr. Chun-Chieh Wu of National Taiwan University (NTU) established an aircraft reconnaissance program called DOTSTAR. While DOTSTAR aircraft observations have generally improved typhoon forecasts, it remains unclear which types of aircraft observations produce the most significant improvements. In this study conducted in collaboration with Dr. Wu, experimental forecasts incorporating aircraft observations of temperature, humidity, or wind speed will be used to evaluate the relative forecast impact of including each individual type of observation. Better understanding how different aircraft observations affect typhoon forecasts will help optimize future observing missions both in the Pacific and Atlantic basins. Aircraft observations of tropical cyclones are expensive, and it is therefore beneficial to optimize such missions by better understanding which observations improve forecasts the most. The proposed study will use DOTSTAR dropsonde observations to investigate the relative impact of thermodynamic versus dynamic in situ observations on typhoon structure and intensity forecasts. A series of simulations will be performed using the Weather Research and Forecast (WRF) model with a modern data assimilation system to independently assimilate profiles of temperature, moisture, or winds from DOTSTAR dropsondes for a sample of West Pacific typhoons. In addition, a control simulation without dropsonde observations will be performed. Each simulation will be validated against independent observations, global model reanalysis, and best track datasets to evaluate the improvements resulting from assimilating each subset of dropsonde observations. This NSF EAPSI award is funded in collaboration with the National Science Council of Taiwan.