Topic: VP2 Reads Too Cold (Night/AM only)

[abqwx]

I've had my Vantage Pro 2 for 4.5 years.  It now seems that the temperature sensor is reading too cold overnight and eventually winding up 6-10 degrees too cold at the coldest part of the day.  I've just replaced the battery a month or so ago and we get *plenty* of sun (with the station solar panel facing south).

I'm comparing my station to several (4-5) on weatherunderground.com that are consistent with each other and all warmer than mine.  These stations are close enough (spread in all directions within about a 1 mile of my location) to be good for comparisons.

After an hour or two of sun, the station reads accurately compared to the others.

What are the next steps to fix this?

Thanks...

 

[abqwx]

As a quick follow up, it also appears that the humidity sensor reads too high overnight, coming back into line with the other stations during the day.

[north of 60]

only at night?  highly doubt it's a sensor problem,  i live on a hill which is warmer than the lowlands around me at night,  if there was 10 stations in the lowlands and they were all 10 degrees colder than me at night does that mean my sensor is too warm?  not a chance,  cold air sinks to the low elevations. 
you are smoking dope if you think your sensor is only wrong at night and not during daytime,  the sensor measures the temp and has no clue what time it is or if it's night or day.
as well the humidity reading should be higher,  both high and lowlands will have the same dewpoint but since your temp is colder and closer to the dewpoint you humidity reading will be higher.
even though other stations is only a mile away i think your temps are correct for location,  i live on hill,   just 600 feet lower than i am down the road the temp can be 10 degrees or more colder at night,  my and their sensors are correct.

[abqwx]

Perhaps I should have been clearer about the proximity of the sensors to my location- all the sensors are at roughly the same elevation (+- 25feet).  Two of the sensors are just a few blocks away in an urban neighborhood area- I get that temperatures can vary over a small distance- this seems too significant to me given the very close proximity of the other sensors.

For the last three days, the high temperature I've observed has been within 0.5 degrees F of the other stations.

For the last three days, my low temperature has recorded at least six degrees lower than *all* of the other stations in the area. 

I know it sounds strange- if I wasn't at a loss for an explanation, I wouldn't have posted here.

[Bushman]

You sure the battery is good?  Is it installed with correct polarity?  If both are true, I would bring it inside to run some controlled tests.  Or maybe your sensor block is failing?

[dalecoy]

I've had my Vantage Pro 2 for 4.5 years.  It now seems that the temperature sensor is reading too cold overnight and eventually winding up 6-10 degrees too cold at the coldest part of the day. 

Exactly where are you located?  Lat/Long if possible.  Is your data being posted somewhere? 

What software are you using to record (and post?) those readings?

[abqwx]

http://www.wunderground.com/weatherstation/WXDailyHistory.asp?ID=KNMALBUQ104

Data's being uploaded via the WeatherLink software using the Weather Underground plugin.  The VP2 console gives the same readings as the online repo.

Lat: 35.1
Long: -106.6

[mulveyr]

Perhaps I should have been clearer about the proximity of the sensors to my location- all the sensors are at roughly the same elevation (+- 25feet).  Two of the sensors are just a few blocks away in an urban neighborhood area- I get that temperatures can vary over a small distance- this seems too significant to me given the very close proximity of the other sensors.

For the last three days, the high temperature I've observed has been within 0.5 degrees F of the other stations.

For the last three days, my low temperature has recorded at least six degrees lower than *all* of the other stations in the area. 

I know it sounds strange- if I wasn't at a loss for an explanation, I wouldn't have posted here.

Any chance that you can verify it with another thermometer?

Temps really can vary a lot;  I'm in a stream valley, and the temps at my station can vary by 5+ degrees from one end of my backyard to another, simply because some accident of topography causes cooler air to pool at the rear.

[dalecoy]

I'm actually familiar with abqwx' location in Albuquerque.  It's an essentially-flat neighborhood (perhaps 1/2 or 1 mile square) of essentially-identical homes, almost all of them single-story and flat-roofed.  [abqwx, please confirm].  There are 2 or 3 other stations reporting on Weather Underground that are in the same (or adjacent) neighborhood, so conditions should be "identical" at all of them - and another dozen or so within perhaps 1 mile.  The official Albuquerque International Airport station is within 2 miles, and has similar altitude and flat geography.  I can't see anything obvious in terms of terrain or nearby objects (houses, factories, etc.) that would have the effect described.

The suggestion of verifying it with another thermometer (even a very cheap one) is good, of course - but I'm betting this is a "sensor/circuitry" issue.  

abqwx, have you physically examined the temperature/humidity sensor?  Bugs, spider webs, crud, etc.?

[Sigdigit]

I'm on the floor of a valley and my night time temps can easily get 5 to 10 degrees colder than surrounding stations just a few miles away.  My humidity values get WAY higher as well at night.  It's not just elevation, it's topography.  My Vantage Pro is dead on, verified with a thermometer.  Even CWOP QC says that it appears my station is located in a valley and my deviations from surrounding stations are normal.  I would think a battery issue would result in failure to transmit data, not erroneously low values.

[dalecoy]

Well, there are valleys and then there are valleys.  abqwx is clearly located in a valley, but please let us know whether Albuquerque New Mexico is similar to the New Jersey valley that you're located in. 

However, you're certainly correct to suggest an actual physical check at the same location.

What are the next steps to fix this?

I think the next step should be to compare your readings with another thermometer placed as close as possible to your VP2.  The somewhat unique terrain of Albuquerque (for instance, Tijeras canyon drainage and down-slope air movement) could be causing very-local variations that you are seeing.  

Readings around 8 AM should be informative (as well as at other times of the day, of course).  

Your near-neighbor KNMALBUQ107 was showing a similar temperature profile this morning - although I'm not sure how much to trust that station (low dew point last week was MINUS 34.9 degrees).

[abqwx]

Hi All-

Thanks for the suggestions-  I've spent some time cleaning and examining the temp/humidity sensor and certainly didn't see anything that looked problematic (dirt, yes, but nothing substantial).  On the plus side, the heat shield has never been cleaner.

I've currently got the VP2 temp sensor, VP2 console (for the "interior" temperature) and another digital thermometer sitting on a table in the shade.  The VP2 temp sensor is 1.9 degrees cooler than the other two (consistently).  I'm going to leave them out over night and see how things look in the AM.

If the 1.9 degree error is consistent, I suppose I could put in an offset on the console....

[dalecoy]

I've currently got the VP2 temp sensor, VP2 console (for the "interior" temperature) and another digital thermometer sitting on a table in the shade.  The VP2 temp sensor is 1.9 degrees cooler than the other two (consistently).  I'm going to leave them out over night and see how things look in the AM.

That sounds good - assuming the current location of the things is similar to the "regular" location.

If the 1.9 degree error is consistent, I suppose I could put in an offset on the console....

Exactly.  You'll know a lot more in the morning.

[abqwx]

After leaving the sensors (all three) next to each other, I can't verify any problem with my Davis outdoor sensor (at least down to the 40 degrees we hit the night I had them all out.

So, I guess, I'm just in a cold spot.  It *really* seems too cold, given the proximity of the other stations and their relative warmth but if I can't verify an inaccurate thermometer, I'm going to go with it.

[dalecoy]

OK, then that was the correct test (before concluding that something was wrong).

Can we presume that your humidity readings (ISS and console) also were reasonably close?

It's interesting that there was almost no wind.  Down-slope "drainage" is very gentle, and wouldn't show as much wind.  And you undoubtedly got a lot of radiation cooling with the clear atmosphere.

Thanks for letting us know the results.  It's useful for other folks in the future.

[abqwx]

Yes- the humidity readings were within a few percent as well. 

The evening I ran the test, I kept the station and console under a covered patio- they didn't get as cold as when mounted on the pole in the backyard.  I plan on double checking the results with the sensor in its normal location when it looks like there's a still, clear night coming. Thursday night looks good for a retest.