Is this stated somewhere by the project team? It doesn't seem right to me. The WRF is a forecast model and almost by definition won't generate a totally accurate forecast. Any model biases or feedback problems would surely propagate to the next run. This is the old "garbage in, garbage out" problem. Why wouldn't they use observational data from The Weather Channel as input to each run? If bad data is propagated over all 182 or 183 iterations, it's almost certain not to resemble reality at the end.

entity

It is an iterative process so they don't send out the next unit for that patch of Africa until they have received the preceding on. However, they are working over a year behind the actual weather so they can compare the results.

Whether they are adjusting or not has not been promulgated.

Mike

Thank you all for your replies to my post. I consider myself admonished and have reduced my WU store accordingly to days and 1 day extra maximum.
In my defence, I have never seen explanations regarding the effect of a large buffer on project execution before. I would feel that it needs to be a pre requisite understanding before one is allowed to sign up for certain projects.
Up until recently I have had a rather unreliable internet which had the habit of shutting off for several days on end and that is why I kept a large buffer - I have now put that right.

@entity,
the various information related to the project way of run as well as to the requirements are disseminated over multiple messages.
Indeed, you point out a really important aspect of some communication mistakes.
ARP1 and FA@H2 have some particularities in terms of deadline.
ARP1 and MIP1 have some "non-affinities".
Mostly Keith and Kevin usually provide some background information ... from time to time.
Surely, it would be fine to make available a kind of data sheet for each project, summarizing the supporting requirements (machine, Internet connexion, listing the main project characteristics), (non-)affinity with other sciences, etc.
Cheers,
Yves

Here's what I do to format the visual webpage result of a workunit, because we know it will vanish after some time:
(Example taken from https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=324305413)
Firstly, I copy the result lines on the webpage to the 'clipboard' by selecting these lines on that webpage:

ARP1_ 0013477_ 029_ 2-- 	Linux Linuxmint 	Linux Mint 20 [5.4.0-48-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9.1)] 	727 	Valid 	10/16/20 23:47:59 	10/18/20 01:09:39 	16.58 	534.7 / 602.8
ARP1_ 0013477_ 029_ 0-- 	Linux 	4.4.0-87-generic 	727 	Valid 	10/9/20 23:44:55 	10/17/20 15:43:58 	16.39 	590.0 / 602.8
ARP1_ 0013477_ 029_ 1-- 	Linux Ubuntu 	Ubuntu 20.04 LTS [5.4.0-47-generic|libc 2.31 (Ubuntu GLIBC 2.31-0ubuntu9)] 	727 	Valid 	10/9/20 23:42:20 	10/13/20 02:56:08 	22.52 	615.6 / 602.8

Next, I paste the contents of the clipboard to a file:
(on the command line: cat > FILE), (then paste the contents of the clipboard, e.g. by typing CTRL+SHIFT+V or by pressing the middle mouse button), (finally, type ENTER and CTRL+D to close the file).
Then I run on the command line:
The output that appears needs to be copied to the body of the post that you are writing ⇒ Select the output, so it will be copied to the clipboard, then paste the contents of the clipboard to the concept of the post.
In this case that will result in:

Result Name          OS           AVN Status Sent Time         Due / Return Time CPUh  Claimed/Granted
ARP1_0013477_029_2-- Linuxmint    727 Valid  10/16/20 23:47:59 10/18/20 01:09:39 16.58   534.7/602.8
ARP1_0013477_029_0-- Linux        727 Valid   10/9/20 23:44:55 10/17/20 15:43:58 16.39   590.0/602.8
ARP1_0013477_029_1-- Linux Ubuntu 727 Valid   10/9/20 23:42:20 10/13/20 02:56:08 22.52   615.6/602.8

[Copied from Workunit Status, generated by wcgformat (using these options: -f -o)]

If it so happens that the lines are still too long (after clicking 'Preview'), add another 'f' to the options of wcgformat:and try again. In this case the output in the post will result in:

Result Name          OS           AVN Status Sent Time         Due / Return Time CPUh  Claimed/Granted
ARP1_0013477_029_2-- Linuxmint    727 Valid  10/16/20 23:47:59 10/18/20 01:09:39 16.58   534.7/602.8
ARP1_0013477_029_0-- Linux        727 Valid   10/9/20 23:44:55 10/17/20 15:43:58 16.39   590.0/602.8
ARP1_0013477_029_1-- Linux Ubuntu 727 Valid   10/9/20 23:42:20 10/13/20 02:56:08 22.52   615.6/602.8

[Copied from Workunit Status, generated by wcgformat (using these options: -ff -o)]
As you can see, that wasn't needed in this case.