How come, when temperature can be measured to an accuracy of 0.1C on
instruments costing under £100, almost all RH instruments quote an
accuracy of +/- 5% at best?
--
Paul Hyett, Cheltenham (change 'invalid83261' to 'blueyonder' to email me)

On 23 Sep, 17:35, Paul Hyett wrote:
How come, when temperature can be measured to an accuracy of 0.1C on
instruments costing under £100, almost all RH instruments quote an
accuracy of +/- 5% at best?

Hi Paul

At Bablake, we have a brand new Met Office AWS, we have a Campbell
Scientific AWS still maintained to Met Office standards and a
conventional Met Office screen with calibrated, Met Office checked
sheathed wet & dry bulbs from which we work out RH each day.

The AWSs have RH electronic sensors, but they don't always agree with
each other, and they can both be several % points at variance with the
wet'dry bulb thermometers. So there you have it - 3 readings every day
and all can be different even in their screens!

However, I don't use a whirling psychrometer very often, but if I did
I thinkk that might be a fourth different reading!

Does that help?

Thought not:-)

Steve Jackson FRMetS
Bablake Weather Station
Coventry UK
www.bablakeweather.co.uk

On Sep 23, 7:19*pm, Steve J wrote:
On 23 Sep, 17:35, Paul Hyett wrote:

How come, when temperature can be measured to an accuracy of 0.1C on
instruments costing under £100, almost all RH instruments quote an
accuracy of +/- 5% at best?

Hi Paul

At Bablake, we have a brand new Met Office AWS, we have a Campbell
Scientific AWS still maintained to Met Office standards and a
conventional Met Office screen with calibrated, Met Office checked
sheathed wet & dry bulbs from which we work out RH each day.

The AWSs have RH electronic sensors, but they don't always agree with
each other, and they can both be several % points at variance with the
wet'dry bulb thermometers. So there you have it - 3 readings every day
and all can be different even in their screens!

However, I don't use a whirling psychrometer very often, but if I did
I thinkk that might be a fourth different reading!

Does that help?

Thought not:-)

Steve Jackson FRMetS
Bablake Weather Station
Coventry UKwww.bablakeweather.co.uk

The humidity measured from wet- and dry-bulbs in a screen is
unreliable because of the unknown speed of the draught over the
bulbs. The formula used is something of a compromise, a guess even,
at the speed of the air movement over the bulbs and at low wind speeds
the wet-bulb depression is strongly dependent on the wind speed. A
whirling psychrometer is much more reliable (if whirled quick enough)
and is the method I use though I don't do it on a regular basis. I
don't know what type of sensor is used in AWS's - it probably doesn't
involve the evaporation of water but depends on the variation of
properties of some substance with humidity and is therefore likely to
be no better than a hair hygrometer and may have a time lag and also
respond differently at different temperatures. If I were still at
work I'd suck 100 litres of air through a cold trap and weigh the
water - not feasible, obviously. It's quite a tricky problem, it
seems, but ought not to be.

Tudor Hughes, Warlingham, Surrey

On Wed, 23 Sep 2009 at 11:19:54, Steve J
wrote in uk.sci.weather :

The AWSs have RH electronic sensors, but they don't always agree with
each other

So I've noticed with my own instruments - and they seem to have especial
difficulty when the RH is very high.
--
Paul Hyett, Cheltenham (change 'invalid83261' to 'blueyonder' to email me)

On 23 Sep, 19:19, Steve J wrote:
On 23 Sep, 17:35, Paul Hyett wrote:

How come, when temperature can be measured to an accuracy of 0.1C on
instruments costing under £100, almost all RH instruments quote an
accuracy of +/- 5% at best?

At Bablake, we have a brand new Met Office AWS, we have a Campbell
Scientific AWS still maintained to Met Office standards and a
conventional Met Office screen with calibrated, Met Office checked
sheathed wet & dry bulbs from which we work out RH each day.

The AWSs have RH electronic sensors, but they don't always agree with
each other, and they can both be several % points at variance with the
wet'dry bulb thermometers. So there you have it - 3 readings every day
and all can be different even in their screens!


Steve Jackson FRMetS
Bablake Weather Station
Coventry UKwww.bablakeweather.co.uk

Here at Campbell Scientific we offer 4 combined Temp & RH probes at
varying cost and with varying accuracys (BTW the AWS we sell are
usually custom configured with sensors, datalogger, comms, and so on
all supplied to required specification using either our own sensors or
leading models from other manufacturers).

All the temp & RH probes we sell have better accuracy than the
original post suggested. For example our own sensor, the CS215, offers
accuracy of ±2% at 25° between 10-90% RH and ±4% between 0 &10% and
between 90 & 100%. The Rotronics MP100A we also sell offers ±1%
between 5 & 95% and ±2% 5% and 95%. Response times and hysteresis
also vary from model to model. All such sensors are factory calibrated
in a controlled humidity environment - in the case of the CS215 they
are calibrated to NPL and NIST standards. Of course accuracy comes at
a price and all 4 models we sell are well above the £100 figure
originally mentioned.

Spec sheets for all the probes can be found he
http://www.campbellsci.co.uk/index.cfm?id=463 - prices available upon
request.

Iain Thornton
Campbell Scientific Europe
www.campbellsci.eu