date: Tue, 17 Jan 2006 16:23:07 -0000 from: "Makin, Janet" subject: PRUDENCE range of results to: gspraggs@anglianwater.co.uk, p.jones@uea.ac.uk, sdw@hrwallingford.co.uk, t.osborn@uea.ac.uk, "EA - Rob Wilby" Dear all, Yes I agree too. Is there any merit though in the WRONG-HadRM3H as Tim found it lies in the middle of the range for the Ouse, while for the Eden it is in the middle during the summer drying, but at the top of the range for the winter wetting and thinks we could interpret the WRONG-HadRM3H results, including LoS results, as being a typical central PRUDENCE scenario? Janet Janet Makin Hydrology Manager United Utilities Water PLC Water Supply-Demand Team, Thirlmere House Ground Floor, Lingley Mere, Great Sankey, Warrington, WA5 3LP Tel: 01925 537060 Fax: 01925 464766 Email: janet.makin@uuplc.co.uk -----Original Message----- From: Makin, Janet Sent: 17 January 2006 15:16 To: Makin, Janet Subject: PRUDENCE range of results -----Original Message----- From: Spraggs Gerry E [[1]mailto:gSpraggs@anglianwater.co.uk] Sent: 16 January 2006 16:29 To: Rob Wilby; s.wade@hrwallingford.co.uk; p.jones@uea.ac.uk; t.osborn@uea.ac.uk; Makin, Janet Subject: RE: PRUDENCE range of results Dear All, I agree with Rob that, as I suggested at the meeting we should discard the earlier results. I think it would over-complicate the report to try and include the wrong results as sensitivity runs and the time will be better spent focussing on interpreting the correct ones. I will endeavour to fit in some runs with Tim's new data during the next couple of weeks, possibly concentrating on Levels of Service yield because Steve's model is less developed in that area. Best Regards Gerry -----Original Message----- From: Rob Wilby [[2]mailto:rob.wilby@environment-agency.gov.uk] Sent: 16 January 2006 16:09 To: Spraggs Gerry E; s.wade@hrwallingford.co.uk; p.jones@uea.ac.uk; t.osborn@uea.ac.uk; Janet.Makin@uuplc.co.uk Subject: Re: PRUDENCE range of results Dear All Looking at Tim's plots, I'm now of the view that we simply discard the earlier results rather than complicate life. Gerry and Janet's historic runs are clearly still of immense value for assessing the systems' behaviour over the last 200-yrs and for checking the realism of Steven's 'emulator'. The case for analysing the future runs using the spreadsheet method is readily made in terms of: 1) consistency between regions; 2) speed/efficiency for multiple runs and 3) potentially, sensitivity testing. How does everyone else feel about this. Tim - assuming that everyone is content with the above, it would be good to have an amendment to your plots removing the three wrong runs. Your expanded set of GCM-RCM combinations certainly helps place our scenarios in a broader context. Cheers, Rob >>> Tim Osborn 01/16/06 02:45pm >>> Dear Rob, Janet, Gerry, Steven and Phil, please find attached a graphic showing the range of precipitation results from the PRUDENCE set of regional climate model simulations. In addition to the 3 wrong scenarios, plus my 3 corrections that I'd already completed, I downloaded and analyzed results from a further 12 GCM-RCM combinations - please see the key, with the driving GCM listed first, followed by the RCM and the ensemble member (1 in all cases, except for HadAM3H-HadRM3H-M3, where the M3 indicates that I used the mean of the 3 ensembles members used for the UKCIP02 scenarios, rather than the PRUDENCE data). The top three in the list are the correct data that we should be using. They do span the PRUDENCE range fairly well (the range of all correct RCMs is shaded pale orange), as originally planned (e.g., in summer, ARPEGE (red) is near the top of the range with only moderate decreases in precipitation, and HadRM3H (dark purple) is near to the bottom of the range with strong decreases). The wrong data (black) lie partly within the PRUDENCE range. WRONG-HadRM3H (diamonds) lies in the middle of the range for the Ouse, while for the Eden it is in the middle during the summer drying, but at the top of the range for the winter wetting. I think we could interpret the the WRONG-HadRM3H results, including LoS results, as being a typical central PRUDENCE scenario. WRONG-HIRHAM (squares) tend to be near the top of the PRUDENCE range (wetting or less drying), though the only critical departure is the huge increase in December precipitation for the Ouse. The WRONG-HIRHAM results could only be interpreted as a sensitivity case, outside the PRUDENCE range, but included in the analysis as an example of stronger winter increases in precipitation than simulated by the available RCMs. For the Ouse, the fractional change in precipitation averaged over October-March is 1.41 (i.e. 41% increase - the huge December value is ameliorated in the 6-month winter mean). It seems valid to include a sensitivity case like this - a 41% increase in winter precip by the 2080s under a fairly high emissions scenario cannot be discounted - the old NIES GCM shown in UKCIP02 had a >60% increase for this case, see fig 28 of UKCIP02. WRONG-ARPEGE (triangles) are clearly outside the PRUDENCE range in summer, especially for the Ouse. The WRONG-ARPEGE results could only be included if interpreted as a clear departure from the range of possibilities indicated by the PRUDENCE RCMs. Note that the weak winter increases for the Ouse (below the PRUDENCE range in December) might explain why yields weren't found to increase as much as the increased summer rainfall might have led us to expect. [Sorry to focus on the Ouse in my discussion above - it's just that the errors are larger there.] I hope this figure proves useful to the steering group in deciding how to deal with the results from the three wrong scenarios. Comments anyone? Best regards Tim << File: prud_prec.gif >> =========================================================================================== ============================= The information contained in this e-mail is intended only for the individual to whom it is addressed. It may contain legally privileged or confidential information or otherwise be exempt from disclosure. If you have received this message in error or there are any problems, please notify the sender immediately and delete the message from your computer. You must not use, disclose, copy or alter this message for any unauthorised purpose. Neither United Utilities PLC nor any of its subsidiaries will be liable for any direct, special, indirect or consequential damages as a result of any virus being passed on, or arising from the alteration of the contents of this message by a third party. United Utilities PLC (England and Wales No.2366616) registered office: Dawson House, Great Sankey, Warrington, WA5 3LW. =========================================================================================== =============================