|
|
发表于 2007-8-11 15:46:46
|
显示全部楼层
原帖由 YsMilan 于 2007-8-1 11:12 发表 
计算程序是:
UK Met Office HadSM3 Slab Model 5.06
任务是:
hadsm3fub_0110_005896470_6
以前没见过,预计完成时间只要464小时,是什么东东?
Met office 官網：http://www.metoffice.gov.uk/
http://www.metoffice.gov.uk/rese ... ette/sept01/um.html <--HadSM3 相關說明 節錄如下：
Figure 3. Global mean temperatures from some early test runs of the PC (32-bit) version of HadSM3.
Preliminary results from three runs are shown in Fig. 3. The solid lines show 10-year calibration runs, for which period oceanic heat flux convergence fields are computed. The dashed lines show subsequent control integrations with present-day CO2 levels, while the dotted lines show perturbed integrations in which CO2 levels are doubled. Each line shade corresponds to a different setting of various parameters in the model cloud and precipitation schemes, all within ranges of uncertainty regarding the processes these parameters represent.
These runs are not in equilibrium after only 10 years of integration: the objective of these experiments is to identify parameter combinations that have a large impact on sensitivity. Runs involving these combinations will then be extended a further 10 years to establish their sensitivity. Even with parameter settings well within current uncertainty ranges, model behaviour can be quite unexpected: one of the 2 x CO2 runs has frozen over after only a couple of years!
The next stage will be to expand this to an 80-member ensemble — using a Linux-based release of the PUM — within the academic community. Later in the year, a Windows version will be released to the general public. (Although further development is needed, the PUM has already been successfully run under Windows.) Following on from these slab model ensembles, there will be a second experiment using the full coupled model — HadCM3L — to simulate the period 1950–2050. Preparatory spin-up runs are already under way and it is hoped that we can achieve an ensemble size of over a million at that stage. This is a very reasonable proposal, given that 17,000 people have already signed up on the climateprediction.com web site.
原文資料如上： Google 翻譯工具結果如下：
初步結果,從三個班次列圖. 3 .實線顯示, 10年期校準運行時,該時期海洋熱流場銜接計算.該虛線顯示以後控制結合當前天的二氧化碳含量,而虛線顯示攝整合中的二氧化碳含量增加一倍.每條生產線綠蔭相當於一個不同的設置各種模型中的參數雲和降水的計劃,所有範圍內的不確定性對於這些過程參數代表.
這些班次不均衡後,只有10多年的一體化:客觀這些實驗是為了找出參數組合,有非常大的影響的敏感性.運行涉及這些組合則將擴大再10年,以確定其敏感性.即使參數設置好當前不確定範圍,行為模式,可以很意外:一個2 ×二氧化碳運行已凍結後,僅僅幾年!
下一階段將擴大到80名成員合奏-使用一個基於linux釋放的轉折點--學術團體.今年晚些時候,視窗版將公佈給大眾. (雖然需要進一步發展,陶瓷已成功運行windows下) .以下就從這些板組合模型,將有第二次試驗充分利用耦合模型hadcm3l -模擬1950-2050年期間.籌備自旋- 14秒251已經在進行之中,這是我們希望能達到一個合奏人數超過100萬,在那個階段.這是一個很合理的建議,鑑於17000人,已簽訂了關於climateprediction.com網站.
以原文文意為主~~~參考囉~~~
[ 本帖最后由 alexpon 于 2007-8-11 15:48 编辑 ] |
|
