[wxqc] Dew Point errors

Sat Aug 6 15:22:14 EDT 2005

Hi Jim,

The entire mid-Atlantic, including Middletown, VA, has been bathed in a 
maritime tropical (MT) airmass for several weeks.  MT airmasses are 
characterized by high dew points, typically mid 60's to low 70's F, 
which is what your adjacent stations have been reporting.  I am sure you 
are correct when you say other area stations receive more rainfall than 
Middletown due to orographic (mountain) influences, but dew point 
measurements are not exactly the same thing as high high dew points 
don't always mean high precip and low dew points don't always mean low 
precip.  Fact is CW1500 and KOKV have just about the same elevation 
above sea level as your station (e.g. low 700 ft).

Rooftop siting will influence your temperature measurements, as it is 
possible the tar shingles will had to the ambient air temperature during 
summer afternoons.  This is why it is suggested that temperature and 
humidity measurements be made at 5 ft above ground level on a "natural" 
and representative ground cover (grass, if grass is typical for your 
area; not concrete other man-made surfaces).   In almost all cases, the 
temperature and humidity sensors are on the same circuit board and must 
be co-located.

Humidity measurements will not suffer terribly from a non-standard 
vertical siting of 10, 15, 20, or 25 ft above ground as humidity, in a 
well mixed atmosphere, does not change dramatically with elevation over 
the short distances we are talking about.  However, artificial sources 
of moisture like sprinkler systems can add very localized and 
unrepresentative moisture to the air you are sampling and cause a 
humidity measurement to be too high.  Both Philip and I have small ponds 
near our humidity sensors (me:  neighbor's standard sized pool about 10 
ft north, Philip:  frog pond a few feet away from his humidity sensor).  
We have pretty much concluded that these relatively small water sources 
do not add enough water vapor to the air to make a difference in our day 
to day humidity measurements.  However, I have seen some really violent 
daytime failures of some dew point sensors, typically those that are 
not  shielded from the sun.  In this failure mode, the dew 
point/humidity drops dramatically during the daylight hours and recovers 
at night.  The predominant problem with humidity measurements is one of 
calibration, typically they run too high, but sometimes too low, as I 
suspect your humidity sensor is doing. 

I suggest doing nothing right now, lets just watch the measurements for 
a few weeks to get a better idea of your sensor's performance under a 
variety of moisture ranges (one day we WILL see a nice cold front).  
Your measurement are really not too fat off from the analysis in most 
cases at any rate.

Hope this helps,

Dave

Jim Wahler wrote:

>Hi Dave,
>
>It's not a vary good comparison.  I live in a 'rain shadow' created by the
>Massanutten Mtns.  KOKV and C1500 both get much more rain than I do.  Plus
>my station is on a tripod on my roof abbout 25' above the ground.
>Do you think I should enter an offset in my VWS program?
>
>Jim
>
>----- Original Message ----- 
>From: "Dave Helms" <dshelms at comcast.net>
>To: "Discussion of data quality issues" <wxqc at lists.gladstonefamily.net>
>Sent: Saturday, 06 August, 2005 13:03
>Subject: Re: [wxqc] Dew Point errors
>
>
>  
>
>>Hi Jim,
>>
>>Your dew point is about 3 F degrees too low (measurements too dry) as
>>compared to the analysis (e.g. your adjacent stations).  See here in
>>comparison to the Winchester Airport (KOKV) observation:
>>
>>    
>>
>http://weather.gladstonefamily.net/qchart/C3977?date=20050806&addnl=KOKV&Add+to+charts=Add+to+charts&.cgifields=addnl
>  
>
>>The message says that you need to <add> 7 F degrees to your dew point an
>>that day to bring it up to the humidity of the nearby stations.
>>However, the 7F degree was just a single day's sample.  I suspect your
>>are running 3-4F too dry most of the time.
>>
>>
>>Hope this helps,
>>
>>Dave
>>CW0351
>>
>>P.S.  You still need to update your position in your VWSAPRS software so
>>FINDU plots your station in "downtown" Middletown, as per our previous
>>email.
>>
>>Jim Wahler wrote:
>>
>>    
>>
>>>*I recently began sending data from my WM918 to APRS, after about 15
>>>months of uploading to Wunderground.  Yesterday I received a message
>>>from wxqc listing a series of 'Observerd Errors' in the Dew Point data
>>>to APRS/CWOP.  It indicated that my readings were about 7F too high.
>>>On checking the graphs I found that I was actually below the analysis,
>>>as well as the other stations near me.  I thought that I needed to
>>>adjust my Dew Point, but now I'm confused about how, which direction,
>>>and IF I really need to do anything at all.*
>>>**
>>>*Thanks for your help & suggestions.*
>>>*Jim - C3977*
>>>
>>>------------------------------------------------------------------------
>>>
>>>_______________________________________________
>>>wxqc mailing list
>>>wxqc at lists.gladstonefamily.net
>>>http://pond1.gladstonefamily.net:8080/mailman/listinfo/wxqc
>>>
>>>The contents of this message are the responsibility of the author.
>>>
>>>      
>>>
>>_______________________________________________
>>wxqc mailing list
>>wxqc at lists.gladstonefamily.net
>>http://pond1.gladstonefamily.net:8080/mailman/listinfo/wxqc
>>
>>The contents of this message are the responsibility of the author.
>>    
>>
>
>_______________________________________________
>wxqc mailing list
>wxqc at lists.gladstonefamily.net
>http://pond1.gladstonefamily.net:8080/mailman/listinfo/wxqc
>
>The contents of this message are the responsibility of the author.
>
>
>  
>