CWOP Red X

[Felix]

Do you live in fear of getting a CWOP Red X?

My friend and fellow weather nut who lives a few miles away is on vacation and his station has gone haywire in his absence...to the point where CWOP is giving him multiple Red X grades for significant differences between his reported vs CWOP predictive values.

His barometric pressure readings were drifting even before he left on holiday so I wasn't all that surprised when he got the dreaded Red X followed a few days later by a Level III Spatial Consistency notice (suspected sensor failure).

Plus, his vigorous lawn sprinkling is affecting the station RH measurements so he's getting another Red X for Dewpoint. Widely varying dewpoint measurements vs CWOP predicted values are showing up as unexpectedly large standard deviation calculations which is putting the station way out of the box on the Quality Spread Distribution Page.

But he's an easy going chap, unflappable and unconcerned. It'll all be fixed in good time when he returns. In fact I rather like his attitude, "It's only a hobby and CWOP sure isn't going to ruin my vacation."   

On a related note, I sometimes wonder just how reliable those CWOP predictive algorithms are anyway. For example, a nearby airport is currently well out of the box (Quality Spread Distribution Page) for wind speed! Now if there is one thing I'd expect the airport to have right, it would sure be the takeoff/landing winds, temps and pressure. So when the CWOP predictive values are significantly different from the airport actuals, I raise the BS flag about their algorithms. I've looked at their station and the siting parameters exceed every guideline I've ever read and sensors are periodically logged and calibrated according to FAA guidelines.

[Blicj11]

Felix:

You raise a good point. I have been living in CWOP Red X land for several months now. It does not always necessarily mean that your sensors are wrong. Depending on where in the world you are and who else is in the area, someone else's sensors could be faulty. And CWOP does not reveal which stations they use for comparison so it is a wee bit frustrating to know if it is your unit or not.

In my case, the local airport definitely reports out of the box barometric pressure. However, after some investigation, I eventually decided that the CWOP Red X was my earliest indication that my pressure sensor is slowly drifting away from accurate readings. My unit is 7 years old and now has two sensors on their way down.

After weighing all the alternatives, I recently decided to purchase a new Davis Vantage Pro2 Plus. After I get it installed, I will have my old unit refurbished by Davis for US$140 and hold on to it as a spare.

[elagache]

Dear Felix, Blick, and WeatherCat fans,

Do you live in fear of getting a CWOP Red X?

I suppose the answer depends on how much you want to contribute to the network of "reliable" data.  If that is a very important goal for you, then I suppose you need to sweat out all the details as to why the CWOP quality assessment tests are failing.

In my case, my station has been given a "red X" for temperature for as long as it has been up.  It's still there, see for yourself:

http://weather.gladstonefamily.net/site/d3835

However, I know this to be an error on the part of the CWOP assessment.  I have two Davis VP-2 thermometers connected to my station.  When the sun isn't beating on the one without the radiation shield, they agree within a degree most of the time.  My suspicion is that our house is unusually cool for the region because we have so many nearby trees.  However, my goal isn't to collect data for the world, it is to know the temperature is in - my own digs!

I assume it means that my CWOP data isn't used for temperature and that's fine.  I live in a moderately dense location.  I assume anyone interested in the weather around here has a decent idea of what is actually going on.  I cannot collect good wind data without putting up a 100 foot tower, so what I do here is always a compromise.

My assumption is that any researcher using CWOP data applies some advanced statistical techniques to ferret out bad data while extracting overall trends.  When I was involved in the "Great American fish count" during my scuba diving days, I was told there were ways for researchers to compensate for errors amateur divers made in the fish count.  Similar techniques can be used for CWOP data.

So for me there are no worries in "da' dreaded CWOP Red X".

Cheers, Edouard 

[Randall75]

Hi cat Family
 The problem with the Gladstone Monitoring is it takes in affect all the weather stations around you so yours might be OK but your neighbors could cause you to get and red X
I came up with an script to check the 3 Airports that are with in 20 miles of my house and I am within 1 or 2 numbers of them, most of the time I am exactly with 2 of them so Don't always go by the reading you get there go by the Madis readings the 1 or 2 thumbs up these are more acturate than the red X's
Here is a graph comparison of gladstone's reports
http://weather.gladstonefamily.net/qchart/C6734?date=&addnl=KVTA&addnl=KZZV
here is shows my Barometer reading low but if I check with the same 3 airports I am the same as there's so something is not getting calculated right on his pages


cheers


 

[Felix]

...go by the Madis readings the 1 or 2 thumbs up these are more actuate than the red X's

You're certainly not the first who's said that!

Looks like you're slightly out of the box on one parameter and on the line on another as of the time of this post.

http://weather.gladstonefamily.net/cgi-bin/wxqual.pl?site=C6734

I share your frustration with Phil's algorithms as well. The problem is, he apparently doesn't have a way of quickly disregarding bogus data which can skew the averages and the center of mass. Eventually his program figures it out but that takes a month plus. It's just crazy that there's so much difference between the two QC tracking systems.

Would seem to me when a station gets two thumbs up from MADIS, Phil's QC program should require manual intervention to award a Red X...especially if a PWS was tracking closely with one or more close-by NWS sites. I've occasionally seen sites close to me with a Red X and two thumbs up and when I moused over the thumbs, the MADIS rating was 98 or 99% with an "OK" QC. I think maybe Phil's algorithms give too much weighting to the standard deviation.

[Blicj11]

I agree with you, Felix and Randall. In my case, MADIS is giving me two thumbs down, with a QC rating of 34%, which was my first indication that I have a bad sensor. 

[Randall75]

Hi Blicj and WC fans
 Also change it from daily to 3 or 7 days and see what you get if it is better then it is a recent glitch
CWOP wants your barometer reading more than any thing


cheers


 

[Blicj11]

Also change it from daily to 3 or 7 days and see what you get if it is better then it is a recent glitch
CWOP wants your barometer reading more than any thing

Randall, I have previously checked the various settings for the data window and I'm two thumbs down on all of them. Checking my barometer readings manually against other weather stations in my area, including a VP2 very close by is convincing me my sensor has gone bad. i have tried to adjust it on the console and within a couple of weeks it shows readings that don't make a lot of sense.