Toward Seamless Weather-Climate Prediction with a Global High-Resolution Model
Abstract
This is a renewal proposal to extend our previously funded ONR seasonal prediction DRI project. The overall objective of this project is to demonstrate the capability of the HiRAM model framework in reproducing/predicting both short-range weather events such as tropical cyclones (TCs) and long-range climate variability such as Madden-Julian Oscillation (MJO) and El Nino-Southern Oscillation (ENSO). What we have achieved during the past three years are summarized as following. Firstly we systematically evaluated the model performance in simulating the MJO variability and ENSO teleconnection patterns through improved convective scheme. Secondly we conducted extended-range (beyond 7 days) forecast of Hurricane Sandy (2012) in North Atlantic and super typhoon Haiyan (2013) in western North Pacific (WNP). Our results indicate that the genesis of both the super storms can be well predicted at 11-day lead, with a high possibility of detection (POD) rate and a low false alarm rate. The timing of landfall can be also predicted one week ahead for Sandy and two weeks ahead for Haiyan. The predictability sources for the beyond weather scale TC forecast arose from successful prediction of MJO and easterly waves in the model. Although being a case study, this work demonstrates potential capability of beyond weather scale prediction of TCs. The goal of this renewal project is two folds. Firstly, we intend to extend the scope of the aforementioned case study by conducting extended-range (10-30-day) TC predictions for a whole TC season. By doing so, we intend to obtain statistically robust extended-range TC forecast skills. Secondly, by conducting 10-year hindcast experiments, we intend to systematically evaluate the model performance in MJO prediction. For the MJO forecast, the coupled model will be spun up with the observed atmospheric and SST nudging from January 2003 to December 2013. The hindcast will be carried out every 5 days for each month from November to April during 2004-1013. For each forecast, 6 member ensemble runs will be carried out. A detailed diagnosis of the model outputs will follow to reveal the extended-range TC and MJO forecast skills. It is anticipated that through the systematical evaluation of HiRAMÕs performance in long-term simulations and hindcast experiments, the modelÕs capability in predicting high-impact weather events (such as TC) and low-frequency climate modes (such as MJO) will be fully assessed. The outcome of this research project will provide valuable information for constructing a base model framework for the next-generation Navy operational seamless weather-climate forecast system.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 10, 2016
- Source ID
- N000141612260
Entities
People
- Tianming Li
Organizations
- Office of Naval Research
- United States Navy
- University of Hawaiʻi System