ISVHE | IntraSeasonal Variability Hindcast Experiment

Bin Wang1, J.-Y. Lee, H. Hendon, D. Waliser, I.-S. Kang, J. Shukla
and CliPAS ISO Team

1International Pacific Research Center, University of Hawaii, USA

A coordinated Intraseasonal Oscillation (ISO) hindcast experiment, which is supported by APCC, CLIVAR/AAMP, and the AMY (2007-2011), was launched, in January 2009. This project aims to estimate intraseasonal predictability and determine realized prediction skill in the state-of-the-art climate models, including major operational centers’ models worldwide. While the establishment of the Madden-Julian Oscillation (MJO) diagnostic metric and the coordination of operational forecast activity have made a great advance, there is an urgent need to exploit these efforts to full potential and to produce a multi-model ensemble (MME) intraseasonal forecast. To this end, the lead-dependent model climatologies (i.e. multi-decade hindcast data sets) are an intrinsic requirement to properly quantify the skills of each model and to produce MME prediction as a function of lead-time and season. The development and analysis of the multi-model hindcast experiments is a critical step for advancing climate prediction on subseasonal time scale. The ISO hindcast experiment includes a set of retrospective ISO forecasts that covers the last 20 years from 1989 to 2008. The minimum (standard) specification of the hindcast are: (a) Prediction is initiated every 10 days on 1st, 11st, and 21st of each calendar month throughout the entire 20-year period: (b) Integration length for each forecast is 45 days; (c) The number of ensemble for each forecast is at least 5. The free coupled run or AMIP-type run is served as control experiment for better understanding the dependence of the prediction on initial conditions and better defining metrics that measure the “drift” of the model toward their intrinsic MJO/Monsoon ISO modes. Currently, 11 modeling groups have finished their experiments. The current status of MJO and Monsoon ISO hindcast capacity will be discussed using seven coupled models and four atmospheric-alone models in terms of MJO diagnostics developed by CLIVAR/MJO working group and Monsoon ISO evaluation metrics developed by CliPAS ISO team.



Programmatic: MJO Task Force, CLIVAR AAMP, Asian Monsoon Year (AMY)
Financial: Asian Pacific Climate Center, NOAA Climate Program Office and IPRC