Anyone successfully filter a report/parameter?

This is the place to get help with Digital Atmosphere, suggest improvements, and get tips.
Post Reply
Wally
Member
Posts: 68
Joined: Wed Jun 15, 2011 11:44 am

Anyone successfully filter a report/parameter?

Post by Wally »

Has anyone had any luck lately omitting a station temporarily? Seems like it was possible to even eliminate a parameter (surface pressure in mb) of any station in an exclusion list in a digatmos.rul file. The offending station is K1M4 in NW Alabama who is always reporting SLP around 965 mb, and it really is messing up analyses!
I could remove the station, but I like the rest of the data, and it would take finesse to remove and put back in master station file.

Thanks.
geophi
Member
Posts: 46
Joined: Thu Jul 15, 2004 3:42 pm
Location: Kansas

Re: Anyone successfully filter a report/parameter?

Post by geophi »

Wally wrote:Has anyone had any luck lately omitting a station temporarily? Seems like it was possible to even eliminate a parameter (surface pressure in mb) of any station in an exclusion list in a digatmos.rul file. The offending station is K1M4 in NW Alabama who is always reporting SLP around 965 mb, and it really is messing up analyses!
I could remove the station, but I like the rest of the data, and it would take finesse to remove and put back in master station file.

Thanks.
I'm having the same problem with KSQI and SLP. I don't know if it has anything to do with the fact I upgraded DA 2.09 Pro to this DA 3.20 Pro on the same workstation without uninstalling? I really should try a clean install on another PC to see if blacklisting a parameter from a site, or a site completely, would work that way. The upgrade install was quite messy if you ask me.
texsky
Senior Member
Posts: 172
Joined: Fri Dec 12, 2003 9:42 pm

Re: Anyone successfully filter a report/parameter?

Post by texsky »

I've had the same issues Wally. Deleting the entire station was the only fix. Tim is scheduled to start some DA programming soon, maybe he'll address this one.

Bob
geophi
Member
Posts: 46
Joined: Thu Jul 15, 2004 3:42 pm
Location: Kansas

Re: Anyone successfully filter a report/parameter?

Post by geophi »

Has anyone brought it to his attention through e-mail? I'm not sure he ever checks the forum, except very near a new release.
geophi
Member
Posts: 46
Joined: Thu Jul 15, 2004 3:42 pm
Location: Kansas

Re: Anyone successfully filter a report/parameter?

Post by geophi »

I sent the following to Tim's e-mail address:
Hi Tim,

Running DA Professional 3.20, I am unable to blacklist observation elements when editing and saving station rules to QC out bad data permanently. A good example is KSQI, Sterling, IL which has had bad SLP for a long time. Others have noted the same problem with the blacklist not working in station rules in a forum thread here. viewtopic.php?f=5&t=2232

Below is an example for KSQI, showing that the entry was saved in the digatmos.rul file, but also the plot (and analysis for) the bad SLP that was still decoded, plotted and analyzed. I exited DA after saving the KSQI rule, re-requested the metar data file, and this was the result.

Hopefully this can be fixed. It worked perfectly in 2.09 to edit in the rules. The only way to get rid of the problem, and not have to manually QC it each time, is to delete KSQI from the station tables.

Thanks.

George Phillips
ObsBlacklistError.jpg
ObsBlacklistError.jpg (137.33 KiB) Viewed 25961 times
geophi
Member
Posts: 46
Joined: Thu Jul 15, 2004 3:42 pm
Location: Kansas

Re: Anyone successfully filter a report/parameter?

Post by geophi »

Tim e-mailed me back and said this would be fixed in the next version. Not sure when that will be, but it's on his radar now.
lrdheat
Member
Posts: 38
Joined: Tue Oct 30, 2007 9:08 pm

Re: Anyone successfully filter a report/parameter?

Post by lrdheat »

KXWA,KSDY,KEKY,KRYY also have bad MLSP's. KPTV needs to be relocated to it's California location. CWSI needs to be moved from it's near Ottawa location to it's location northeast of Quebec City. In the mean time, using quality control, and going to those stations and putting -9999 into the SLP cell (and deleting repeat obs or earlier obs from those stations works, is an inconvenience, but does not take too much work). KSQI seems to have gone to mostly missing data...wonder if SLP will be addressed when it comes back up. Anyone know where KISN has gone (call letter change?) KMQY, just outside of KBNA, seems to read ~1 mb too high.
geophi
Member
Posts: 46
Joined: Thu Jul 15, 2004 3:42 pm
Location: Kansas

Re: Anyone successfully filter a report/parameter?

Post by geophi »

lrdheat wrote:KXWA,KSDY,KEKY,KRYY also have bad MLSP's. KPTV needs to be relocated to it's California location. CWSI needs to be moved from it's near Ottawa location to it's location northeast of Quebec City. In the mean time, using quality control, and going to those stations and putting -9999 into the SLP cell (and deleting repeat obs or earlier obs from those stations works, is an inconvenience, but does not take too much work). KSQI seems to have gone to mostly missing data...wonder if SLP will be addressed when it comes back up. Anyone know where KISN has gone (call letter change?) KMQY, just outside of KBNA, seems to read ~1 mb too high.
KISN is now KXWA. North Dakota & vicinity is a disaster with respect to really bad SLP.
lrdheat
Member
Posts: 38
Joined: Tue Oct 30, 2007 9:08 pm

Re: Anyone successfully filter a report/parameter?

Post by lrdheat »

I find that it doesn't take to long to quality control those stations by drawing a box around the offending stations, putting -9999 in the pressure field, keeping the other data. Be careful...the offending stations often have duplicate reports...I fix the 1st one I encounter, and delete the duplicates. As to KPTV, and CWSI, I type in the correct lat/long into 1st of duplicate reports (36.02,-119.07 135M for KPTV, 47.28/-70.63 719M for CWSI). It's worth the couple of minutes to do so.

On the U.S. scale of things where station density is high, I find that with MADIS, using 94% density makes a mess of some stations with multiple reports in an hour. For some reason, 92% plots without the duplicate reports. Wish this worked at 94%
Post Reply