Mar 03 2008
Two Wrongs Don’t Make a Right… Unfortunately
Or: If there was any doubt I was stupid, there isn’t anymore.
Or: Where’s the bitch and moan department?
Or: Lack of brain foils scarecrow.
Or: Suggestions for the USHCN that would make it more difficult for them, but easier for me - and that’s all I care about right now.
Any other suggestions for alternative titles for this post?
For whatever reason, I was looking at the station at Troy Lock and Dam. Notice the three points way to the left of the rest? That’s part one of the problem.
The monthly data looks like so:
|
Notice the spikes down to -99.99 during the long duration period on the right? That’s part two of the problem.
Notice the large gap? That’s the third part of the problem.
All that remains is for someone stupid [me] to make a little assumption, and the whole thing turns to crap. For those that haven’t put the pieces together, the time period during the gap doesn’t have missing data values, and I assumed that once the time series started any missing values would have missing values recorded. (Not a totally unreasonable assumption I don’t think.)
Sooo, I’m going to have to do that myself I guess. Since the USHCN is a pain to work with and I can’t post on this topic until I get this issue fixed, I’m going to offer some suggestions for making the data easier to work with (for me at least).
- Binary not ASCII
- Common start year
- Bad values thoughout
- Separate products = separate files
- Incorporate metadata
The only positive thing about an ASCII file is that it’s human-readable. But who opens a 72Mb file to browse through it? A common start year would make it easier to compare time series, although I guess just noting the first entry somehow might save lots of space. 3 and 4 seem obvious, but obviously not. And by incorporating the lat/lon/elev of the station for each entry I wouldn’t have to look elsewhere for it.
This bug should not invalidate previous work that used the FILNET data, since that should have all the bad values filled in already. But I’ll need to check that too I guess.
Is there anything that I’m missing for the B&M department?
(No Ratings Yet)
Loading ...Related Posts:
4 Responses to “Two Wrongs Don’t Make a Right… Unfortunately”
To reduce spam, comments are automatically closed 30 days after the last comment. If you would like to comment on any closed thread, please use the contact form at the top of this page.
[...] 03/Mar/2008: See also Two Wrongs Don’t Make a Right… Unfortunately. First, let’s look at the raw temperature record of a random station in the USHCN. When I [...]
Don’t get your concern. Back up and explain it. You are spazzing out a little.
ATMOZ,
I have an interesting case for you to look at with your method.
in H2001 Hansen mentions that Crater Lake station was deleted in it’s entirety. I spent a couple days looking at it and the surrounding stations and nothing struck me as being odd about this station. This was an utterly subjective assesment so it might be interested to see what you method suggests.
[...] Steven Mosher asks: I have an interesting case for you to look at with your method. [...]