学而时习

7Timer!的诸位用户：

    很抱歉这么晚才宣布这个通知，我应该早一点用中文宣布的。考虑到时间和大家的耐心，我就长话短说了吧。

    经过将近一年的测试，7Timer!小组积累了不少使用GFS数据的经验，其中最重要的就是GFS非常不适合对某个精确地点的预测。GFS(全球预测系统)对很大一片地方的预测非常准确，但对于具体地点的预测则常常出错，所以我花了许多时间回答用户的许多问题，当然其中相当一部分都是抱怨7Timer!的精确度太低的。

    经过考虑，我决定作出以下较大的变动：撤销对某一具体地点的精确预测，改为提供预报图的形式进行预测。也就是说，今后的7Timer!将基本不再提供“面板”一样的预测(当然以前已经申请的将视需要情况处理)，而是都用预测地图的方式来提供预测，这样可以充分发挥GFS数据的长处。目前命令文件都已经写好了，但是网页设计方面还是有一段距离，所以这一段时间大家只能从“预测图”栏目下的中国地图来了解预测情况。当建设好之后，我把全中国划分成5个区域来进行预测。

    在完全实行预测图(或者叫7Timer! V3)制度之前，将有段过渡时期。我将在申请7Timer!那里贴出内容大体和这篇一样的告示，提醒新用户制度的转换。当然以后，如果您有需要，还是可以申请“面板”预测的。

    本文和http://csc.lamost.org/yeiht/wordpress/?p=111 (E文)内容大致相同。

    谢谢各位对7Timer!的支持。

7Timer!工作组

2006.04.28

Hello all,

  I'm sorry -- after one day's working I still could not finish the forecast maps page creating, so we have to experience longer to see the function. The only work published today is the "Very Important Note" page.

  Though I have finish the command writing (it means the ARP could output the images now), but I take a long time to design and try to complete the java tool on the new forecast map page. I failed, however, so we have to wait.

  Now please use the "world-wide map" to see forecast... sorry about this.

Regards,

Quanzhi

Hello all,

  Sorry for this post's delay, I was not able to write it yesterday -- still busy busy and busy.

  I'm now studying the method of seeing. Though I only find a few articles via internet about this topic, I'm still expecting we could have seeing forecast in 7Timer! one day (now I learn it's possible to do that -- the scientists at Mauna Kea successfully use GFS data to produce seeing forecast, I have write them but I haven't receive reply so far).

  But this is not the big change. Now I'm considering if it's necessary to change the operate mode. As you know, now the operate mode is, produce forecast for a pointed position.

  However, after nearly one year's operating, I receive many concern about this kind of mode sometimes go "slightly" wrong. I realize GFS data is good at "global" -- goot at "trend", it's not "Local Forecast System". So, in my consideration, the new mode would be this kind of thing --

  There will not be forecast for any pointed location (however, those which have been existed would be re-consider if it is necessary to keep them running), I will only produce forecast maps for a province, country, etc. Or say, 7Timer! will only predict weather for a large area, instead of local. That would make 7Timer! more useful, I think, because the resolution of GFS data is 1 arc deg, it's waste of 7Timer! to produce "local" forecast.

  So what is going on? Well, in my imagine, firstly we would have a "switch period", at that time I would create some new pages for some big area, and I would encourage new comers to use these "forecast maps" instead of request new 7Timer!s. Before we switch to the new mode, we would experience two months or longer, as now I'm preparing the forthcoming NMT and do not have much time.

  However, I'm now seriously considering this project. If all things go right, I would start it this weekend. If you have any good idea, please let me know -- either mail or post reply here. Thank you!

Best regards,

Quanzhi

Hello all,

  I can annouce the ARP is working now -- all the things are back to normal.

Enjoy!

Quanzhi

Hello all,
After two days' working, I could cheerfully annouce that a new model and a new version of ARP has released.
7Timer! now is using Model-V2e, which is an improve model of previous Model-V2a. I hope it could please everyone of us.
A new ARP also could raise the forecast quality a little, and it could save me time on new request precessing and model upgrading.
I'll stop here now and go on recovering my lessons. A majoy exam will occuer in the following days.
As usual, email me if you have problems/comments/etc.
Best regards,
Quanzhi