Atmospheric Science [A]

 CC:Hall E  Wednesday  1400h

A Meeting of the Models: North American and International Multimodel Forecasting in Research and Operations II Posters

Presiding:  D Collins, NOAA/NWS/NCEP Climate Prediction Center; P Bourgouin, Canadian Meteorological Centre, Environment Canada


TIGGE-based identification of typical weather events in summer over China

* Kan, D (, Chinese National Meteorological Center, No.46 Zchongguancun South Avenue,Haidian district,Beijing,P.R.China, Beijing, 100081, China
Ronghua, J (, Chinese National Meteorological Center, No.46 Zchongguancun South Avenue,Haidian district,Beijing,P.R.China, Beijing, 100081, China

Taking the SOM method as the framework, identification process for the typical weather events in summer over China is designed based on the TIGGE products. The detailed steps are as follows. Firstly, typical weather events over China in summer (June to August) are obtained via the SOM identification to the ERA-40 reanalysis data provided by the ECMWF. Secondly, using the typical events in the first step as the basic patterns, the TIGGE ensemble forecasts can be classified by the SOM method and generate three kinds of products, that is, weather type forecast, probability forecast for weather type and predictability evolution. Thirdly, on the basis of weather type forecasts, the local weather forecast is made through the downscaling analysis of the historical site data. According to the identification process, the 1958-2002 ERA-40 summer data (June to August) are used for the weather type identification of the 500 hPa geopotential height field in East China, with 144 types of typical weather events obtained. Results indicate that, (1) the 144 types can correctly summarize all kinds of the typical weather events in summer over China, such as typhoon, extra-tropical cyclone, the Northeast cold vortex, the westerly trough, west-low-east-high pattern or west-high-east-low pattern, abnormally strong western pacific subtropical high, and so on; (2) stable events, like the westerly trough, display high-frequency, while the transient weather, like typhoon, extra-tropical cyclone and abnormally strong western pacific subtropical high, show relatively low frequency; (3) there is relatively large error in the identification for the transient weather. And then, using the obtained 144 types as the basic patterns, identification of the typical weather events is performed via the SOM method in the context of the 50 ensemble forecasting products from ECMWF since 00:00UTC, 9th July,2007, with 4 kinds of the forecasting products obtained as follows:(1) probability distribution of the weather type at each prediction time, which can help forecasters quickly grasp the future evolution of the weather type; (2) the used case's curve of the predictability against time indicates that within the 84h lead time the predictability decreases quickly with time, but slowly outside the limitation; (3) verifies the weather-type forecasting products , i.e. the weather type of highest probability at each prediction time, show that the SOM method can identify correctly the forecasted typical weather events; (4) the distributions of local downscaling precipitation and anomaly temperature corresponding to forecasting synoptic types, which can help to make fast judgments for the future weather.


Product Development for the multi-model Ensemble Prediction System Based on Ward Analysis

* Tian, W (, National meteorological center, No.46 zhongguancun south anenue,haidian district, beijing, bj 100081, China
Jin, R (, National meteorological center, No.46 zhongguancun south anenue,haidian district, beijing, bj 100081, China

In order to use the products of TIGGE Ensemble Prediction System(EPS), the clustering analysis techniques are introduced, and the Ward analysis is applied to interpret the products of TIGGE EPS. Use this method we can get the main types of the circulations and each one's occurrence probability. But we need pay attention to the system bias, it plays an important parts when combine the forecast from multi-model EPS.The case study for application capacity of EPS clustering products indicates that the Ward analysis can classify the circulation efficiently, and the clustering products are very helpful and convenient for weather forecasters to make use of the EPS information. Till now there is none method can be used to all kinds of circulation, we will use other methods in the further research.


The Use of Ensemble Forecast Systems in MJO Prediction

* Gottschalck, J (, NOAA / Climate Prediction Center, 5200 Auth Rd, Camp Springs, MD , United States
Wheeler, M (, The Centre for Australian Weather and Climate Research, Bureau of Meteorology GPO Box 1289, Melbourne, VIC , Australia
Weickmann, K (, NOAA / ESRL, DSRC, Bldg 33 325 Broadway, Boulder, CO , United States
Waliser, D (, NASA - Jet Propulsion Laboratory, M/S 183-501 4800 Oak Grove Drive, Pasadena, CA , United States
Sperber, K (, Lawrence Livermore National Laboratory, Mail Code L-103 7000 East Avenue, Livermore, CA , United States
Collins, D (, NOAA / Climate Prediction Center, 5200 Auth Rd, Camp Springs, MD , United States

The U.S. Climate Variability and Predictability (CLIVAR) MJO working group (MJOWG) has outlined a strategy to develop and apply a uniform metric for real-time dynamical MJO forecasts. The purpose of this activity is to provide a means to (1) quantitatively compare MJO forecast skill across operational Centres, (2) measure gains in forecast skill over time by a given Centre and the community as a whole, (3) facilitate the development of a multi-model forecast of the MJO, and (4) judge proposed forecast model improvements relative to their performance in representing the MJO. This paper focuses on the use of ensemble forecast systems as part of this project in this new arena. The current status, initial results and future plans of the activity in the context of ensemble information is described. The agreed upon MJO metric is based on extensive deliberations amongst the MJOWG in conjunction with input from a number of operational Centres and follows closely that outlined by Wheeler and Hendon (2004). At the current time, nine operational forecast Centres are or have agreed to participate and include NCEP (US), ECMWF (UK), UKMET (UK), CMC (Canada), ABOM (Australia), CWB (Taiwan), IMD (India), JMA (Japan), and CPTEC (Brazil). The Climate Prediction Center (CPC) within NCEP is hosting the acquisition of the data, application of the MJO metric and realtime display of the standardized forecasts. The activity has already seen application in an operational setting and further standardizing of the forecasts and their illustration along with systematic verification is expected to increase the usefulness and application of these forecasts and lead to more skillful predictions of the MJO and indirectly extra-tropical weather variability influenced by the MJO.