cc: "Tovey Keith Dr \(ENV\)" <K.ToveyatXYZxyz.ac.uk>

date: Wed, 15 Jul 2009 15:07:27 +0100

from: "Metcalfe Peter Mr \(VCO\)" <P.MetcalfeatXYZxyz.ac.uk>

subject: RE:

to: "Jones Philip Prof \(ENV\)" <P.JonesatXYZxyz.ac.uk>

Phil,

Thank you for this.

We need to use the degree days to 'normalise' temperature related energy

demand data for the Government estates' building energy reporting. This

is so the annual heating related energy use can be comparable against

their efficiency targets.

I was concerned that the Oxford (met) data were not so accurate. But if

this difference is likely then it may be that we could risk using this

data.

Obviously an improved dataset for generating monthly degree days over

the coverage of the UK for the last year would be very useful here.

However, I am still unsure whether there is scope, or it is feasible at

all, to interpolate degree days across 5km grids over the UK from BADC

or Met datasets?

This may be something to consider next year if we do indeed succeed in

winning the bid. BRE were the previous incumbents, but I am told they

were expensive.

If interpolations and greater accuracy is possible we may have to pay to

use regional degree days data using a less robust method instead, which

probably doesn't set us apart from other tenders, except perhaps for

writing CRU involvement in an advisory capacity for the initial year or

so.

Regards

Pete

-----------------------------------------------------------------

Peter Metcalfe

School of Environmental Science

University of East Anglia

Norwich

UK

NR4 7TJ

T +44 (0)1603 592583

F +44 (0)1603 591549

-----------------------------------------------------------------

-----Original Message-----

From: P.JonesatXYZxyz.ac.uk [mailto:P.Jones@uea.ac.uk]

Sent: 15 July 2009 14:22

To: Metcalfe Peter Mr (VCO)

Cc: "Goodess Clare Dr \" <c.goodessatXYZxyz.ac.uk>, "Tovey Keith Dr \"

<k.toveyatXYZxyz.ac.uk>, "Jones Philip Prof \" <p.jonesatXYZxyz.ac.uk>

Subject: Re:

Peter,

Clare has now gone to Cuba. I've had a look at the

website.

CRU would get data from BADC. This wouldn't likely

be any more accurate than data from the Oxford site. With

BADC we could calculate numbers for a longer period to

be able to say how unusual recent years have been.

The differences between Norwich and Wattisham relate

to differences in the weather. Oxford are using a consistent

method for all sites. With the fixed thresholds you will

get differences because it could be that Norwich might be warmer

than Wattisham and above a specific threshold.

We could do this with a gridded dataset, but we'd need to

get this from the Met Office. It would be available for

the last 30-40 years.

Cheers

Phil

> Clare,

>

>

>

> I'm nearly there regarding the proposal.

>

>

>

> Hopefully I will get the draft of approaches to you today whilst our

> people here put the contractual and costings together. We aim to send

> off on Friday, or Monday at the very latest.

>

>

>

> I have just a quick couple of questions:

>

>

>

> I was trying to understand how much variation there may be between

> regional and site reported degree days for the same months.

>

>

>

> I did a quick scan of some of the Oxford University degree days data

> online

> (http://www.eci.ox.ac.uk/research/energy/degreedays.php#degreedays)

>

>

>

> Below are data derived from Norwich and Wittering Met sites

>

>

>

> base_15.5

>

> base_15.5

>

> 269.317

>

> 243.789

>

> 277.35

>

> 250.755

>

> 270.6

>

> 220.57

>

> 151.092

>

> 107.064

>

> 136.229

>

> 93.6752

>

> 53.8373

>

> 33.3056

>

> 38.9375

>

> 11.5865

>

> 39.3796

>

> 13.9315

>

> 70.5989

>

> 41.5244

>

> 146.433

>

> 106.199

>

> 255.083

>

> -99.9

>

> -99.9

>

> -99.9

>

>

>

> This is for months Jan to Dec for 2007. There are certainly some

marked

> differences - probably they are not so accurate data since the number

of

> readings per month varied in these (Norwich was av 480 and Wittering

> over 700)

>

>

>

> Oxford's Russell Layberry generates new degree day updates weekly and

> daily. The data is derived from 77 Met. Office weather stations (view

> locations of the stations on the Met Office website

> <http://www.metoffice.gov.uk/education/data/uk_map.html> ) in the UK.

>

>

>

> Clare

>

> 1) Would CRU use of BADC data be more accurate than these - or

can

> you derive more accurate quality processed data using other

techniques?

>

> 2) The possibility of interpolating degree days across the 5km

> grids across the UK sounds really useful for improving the weather

> correction process - but how much extra work is this likely to entail?

>

>

>

> Sorry for the repeat of questions I asked before. I have decided to

> split the proposal into two options depending on the qulity of the

> energy data we have to correct.

>

>

>

> Initially the estimate of ten days minimum included generating degree

> data from hourly temperature data from 20 or so BADC sites with some

> work with 5km grid data inlcuded.

>

>

>

> I want to be sure if I understood correctly that a more sophisticated

> approach involving a 5km grid interpolation of degree days is possible

> over a larger geographical coverage (UK?) from the weather generator

> project, and if so that I have accounted for the time for you to

deliver

> this appropriately and realistically before I commit it to the tender.

>

>

>

> Hope that's ok. I look forward to your response.

>

>

>

> Thank you.

>

>

>

>

>

>

>

> Pete

>

>

>

> -----------------------------------------------------------------

>

> Peter Metcalfe

>

>

>

> School of Environmental Science

>

> University of East Anglia

>

> Norwich

>

> UK

>

> NR4 7TJ

>

> T +44 (0)1603 592583

>

> F +44 (0)1603 591549

>

> -----------------------------------------------------------------

>

>

>

>

## No comments:

## Post a Comment