Transpose-AMIP. Steering committee: Keith Williams (chair), David Williamson, Steve Klein, Christian Jakob, Catherine Senior

Transcription

1 Transpose-AMIP Steering committee: Keith Williams (chair), David Williamson, Steve Klein, Christian Jakob, Catherine Senior WGNE - THORPEX-PDP workshop, Zurich, 08/07/10

2 What is Transpose-AMIP? Basically, running climate models in NWP mode. Joint WGNE-WGCM endorsed activity Core expt for Transpose-AMIP II is to run 64 hindcasts, each 5 days long, initialised from ECMWF YOTC analysis. Optional expt to repeat the same set of hindcasts with NASA MERRA re-analysis or own analysis. The hindcasts are spread through the annual and diurnal cycles during 2008/9 and were chosen to tie in with YOTC and coincide with some of the IOPs in: VOCALS AMY T-PARC Any global modelling centre (NWP or climate) can submit data. Those taking part in CMIP5 should use the same model as is being used for their AMIP simulation.

3 Aims of the project To test model s parametrizations while the circulation is still close to observed. To evaluate processes operating in the model against observations for particular events (e.g. ARM/Cloudnet sites, actual A-train passes, etc.) To compare SCM case study results to full GCM. To be able to comment on the ability of models taking part in CMIP5 to accurately represent fast processes.

4 GFDL precipitation biases Day 3 error AMIP error

5 Example analysis 1: Cloud regimes (MetUM)

6 ISCCP cloud regimes Rossow et al. (2005, GRL)

7 Observed and simulated midlatitude cloud regimes Williams and Webb (2009, Clim Dyn)

8 Cloud regime properties in short-range forecasts Williams and Brooks (2008, J. Clim)

9 Refl. SW Model minus ISCCP DJF surface net down SW bias ISCCP Cl frac Model Cl frac ISCCP Refl. SW Model Refl. SW Warm Southern Ocean due, in part to excessive surface downward SW. This is due to a lack of bright mid-leveltop cloud.

10

11 Comparison with CloudSat using COSP B A A B MetUM: proto-hadgem3 CloudSat

12 Example analysis 2: Dry lower troposphere (NCAR CAM)

13 CAM humidity errors Zonal mean climatological RH bias Hindcast evolution of q biases over SGP site Williamson et al. (2005)

14 Breakdown of hindcast tendencies Williamson et al. (2005)

15 Example analysis 3: SE Pacific Stratocumulus (CAM, GFDL, ECMWF)

16 Investigation of PBL height against EPIC Hannay et al. (2009)

17 Current status of transpose-amip II

18 Received commitments to run Transpose-AMIP II expt from: Met Office (Keith Williams) NCAR (David Williamson) MPI (Bjorn Stevens) IPSL (Sandrine Bony) EC-Earth (Frank Selten) Meteo France (Michel Deque) The first data should become available to download within the next few months.

19 Data formats and the process for downloading will be the same as for CMIP5. The transpose-amip II diagnostic lists are largely based on the CFMIP component of the CMIP5 lists (including COSP output, etc.). Data will be saved globally every 3 hours and the CFMIP sites diagnostic list will be saved every 30 mins. Includes: ARM sites Cloudnet sites GPCI

20 Sounds great! What can I do now? Run the transpose-amip II experiment with your model! MIP tables will be available from the website in the next few weeks and data centres able to receive data in the next few months. Submit a diagnostic subproject summary i.e. a paragraph saying what do you want to do with the data when it becomes available. If there is a lot of interest there is in the data, more modelling centres will conduct the experiment.

21 More information

22 Process-based compositing techniques

23 Examples of evaluation using cloud compositing techniques Williams et al. (2006) Bony et al. (2004)

24 Examples of evaluation using cloud compositing techniques Field and Wood (2006) (Also similar studies by Tselioudis and Jakob, 2002; Norris and Iacobellis, 2005; etc.) Williams and Webb (2009)

25 European circulation regimes

26 Comparing cloud regime error metric with traditional cloud metrics Gleckler et al. (2008) Model A Model B Model C Williams and Webb (2009) Multi-model mean/median Model B Model C Model A HadGEM1

27