Print

Print


Kevin
 
The excessive variability in GACs you observe in repeated runs appears to be an artefact of CLEA UK, not the probabilistic method.  Solutions using Faber Maunsell's CLaRAT software which is a full implementation of CLR10 and subsequent briefing notes 1-4 does not suffer from this problem.
 
Regards,
Jonathan
 
----- Original Message -----
From: [log in to unmask] href="mailto:[log in to unmask]">Kevin Privett
To: [log in to unmask] href="mailto:[log in to unmask]">[log in to unmask]
Sent: Monday, January 15, 2007 4:13 PM
Subject: Variation in CLEA UK output

I though I might share with you a cautionary note concerning the use of CLEA UK.  I have noticed this effect before, and I know others have too, but I am not aware of any discussion of the problem.

 

It concerns the variation in results if you run the same thing several times.  This is to be expected as it is a probabilistic program.  In some cases, the variability is not excessive, but it is in others and it is necessary to carry out many repeated runs before reporting a value.  I have not done enough research to determine what is to blame, or why.  It depends on the chemical and the land use and I have not yet detected a pattern.

 

Consider anthracene as an example.  I have run all 4 land uses at all 3 SOMs to calculate GACs. The land uses with exposure to plant uptake are highly variable, but the others are stable.  There is no real variation with SOM for residential without plant and commercial.  The variation between runs for the other two is so great that it masks any potential variation with SOM, and so I am going to lump them all together.  I have made 5 runs at each SOM or 10 runs at each SOM, but not being a statistician I am not sure if these are enough to be viable. The findings are as follows.

 

Residential without plant uptake varies from 19100 to 20700 mg/kg (mean 19693) – the range is 8.1% of the mean.  No. of runs = 15.

Commercial varies from 434000 to 452000 (mean 440667) – the range is 4.1% of the mean. No. of runs = 15.

Residential with plant uptake varies from 17600 to 57800 (mean 33330) – the range is 120.6% of the mean.  No. of runs = 30.

Allotments varies from 15400 to 83800 (mean 42700) – the range is 160.2% of the mean. No. of runs = 30.

 

As you can see, there is a huge range in the latter two cases.

 

I assume this variability will be removed once CLEA UK is updated to remove the probabilistic functions, as suggested might be the case in The Way Forward.

 

 

 

Regards,

Kevin Privett.

 

Dr Kevin Privett

Geo-Environmental Associate

 

Hydrock Consultants Ltd

Over Court Barns

Over Lane

Almondsbury

Bristol

BS32 4DF

 

Tel: (01454) 619533

Fax: (01454) 614125

[log in to unmask]

Cell phone: (07799) 430870

 

Offices in Bristol, Plymouth, Northampton, Stoke-on-Trent.  www.hydrock.com

 

Disclaimer


The information in this e-mail is confidential and may be read, copied or used only by the intended recipients. If you are not the intended recipient you are hereby notified that any perusal, use, distribution, copying or disclosure is strictly prohibited.  If you have received this e-mail in error please advise us immediately by return e-mail at [log in to unmask]">[log in to unmask] and delete the e-mail document without making a copy. Whilst every effort has been made to ensure this email is virus free, no responsibility is accepted for loss or damage arising from viruses or changes made to this message after it was sent.

 

  Before printing this e-mail please think about the environment