date: Tue Oct 25 15:46:44 2005
from: Keith Briffa <k.briffa@uea.ac.uk>
subject: Review of Proposal  - NE/D001463/1- CONFIDENTIAL
to: paneld@nerc.ac.uk

   Sarah
   here is my assessment of the Huntley proposal . These are good people but I can not in
   conscience rate this as a high proposal in terms of blue skies research or strategic
   outcomes for future  modelling of climate changes. Some of the work is worthy , but none is
   particularly novel and it can not be considered good value for money. I have to ask that
   this review be kept anonymous. Thank you.
   -------------------------------------------------------------------------------------------
   ---------------------------------------------------------
   Your Ref NE/D001463/1
   Principal Investigator Professor B Huntley
   1 No entry
   2 Medium
   3 Whole application
   4
   (a) alpha 3
   What are the proposal's strengths
   The main attribute of the project is the expertise of the proposers in the fields of lake
   sourced Palaeoenvironmental data. Assembling the evidence of the past as a way of testing
   forest dynamics in a vegetation model is good in concept and the selection of the region
   and the types of proxies are as good as could be achieved in most regions. These strengths
   are undermined by limitations in the relevance of the likely outcomes to the proposers
   over-achieving goal.
   What are the proposal's weaknesses
   The over-arching goal of the proposal is stated as to improve knowledge and understanding
   of the rate and character of the response of Boreal and sub-Arctic ecotones to climate
   change, in order to improve parameterisation of seasonal albedo feedbacks in climate change
   models. The extent to which this is achievable depends on the relevance of the results for
   the chosen location to Arctic scale earth system models being used to predict climate
   change.
   The emphasis in the proposal is on one model LandClim which will operate at a spatial
   resolution far beyond the achievable resolution of all climate models being used to project
   climate change in fifty to one hundred years time. Typically these have a resolution of 2.5
   degrees of latitude or about 250km. Even with an embedded high resolution regional model,
   (not routinely run in model projections), this would still likely be 50km or at very best
   25kms. This virtually encompasses the entire range of the proposed study.

   The proposal does not make a convincing case for how the results of this localised
   vegetation modelling would impact on the design of current large modelling of future
   climate or vegetation schemes used by general circulation models. Besides mean climate,
   issues to do with the realistic simulation of extreme short lived or small scale events
   like frost day probability terrain influenced exposure and snow cover, or localised
   palludification must all influence germination and mortality of seedlings and must  be
   parameterised in some way.

   The comparison of different estimates of local climate change produced by a range of
   different lake proxy data is of scientific value and comparing this evidence with
   indications of changing tree cover through time will also produce results that are of
   scientific use. Neither idea can be considered original. For an example see Heinrichs et
   al. in Boreas (2005) pages 192-206 or Lotter et al. in Palaeoclimatology (1997) pages
   395-420, or Heiri et al. (2003) in Palaeo pages 35-53, or Seppa and Birks (2002) in
   Quaternary Research pages 191-199. A review of these and other work shows significant
   variability in even the major subdivisions of inferred Holocene climate or characteristic
   vegetation biomes in the north of Scandinavia. Apparent contradictions in the character and
   timing of transitions are probably the result of different sampling approaches and limited
   sensitivities in the available proxies, coupled with real dating uncertainties.

   Complex ecological changes in lake structure and the chemical and biological evolution of
   internal lake conditions can confound the simple interpretation of different biological
   proxies as monitors of external climate conditions. Climate estimates based on modern day
   training sets for specific proxies typically produce large uncertainty and past climate
   estimates might have uncertainty of an order equal to the magnitude of temperature change
   in this area during most of the Holocene.

   There is also likely to be additional significant uncertainty in the age-depth models for
   different lakes. Even if some tephras are reliably identified they will be likely
   associated with uncertainty of decades to centuries when the stratigraphic uncertainty in a
   particularly lake core sample is taken into account. See Pilcher et al in Boreas (2005)
   pages 136-156. All of the above will limit the accuracy and precision with which climate
   changes can be deduced or can be associated with tree cover changes but it will especially
   limit the degree to which inferences about the phasing and rapidity of tree distribution
   responses to rapid warming.


   The work is scientifically interesting but the reward can be considered medium as far as
   relevance for future climate modelling. The part of the work proposed that deals with
   simulation of present tree distribution using modern climate in the selected region, and
   comparison with satellite and detailed terrain mapping of trees on to GIS seems a useful
   preliminary exercise. This could act as a test of the model sensitivity to climate changes
   in temperature and precipitation that are well enough known over hundreds of years.
   Looking more widely at how well this model performs at different locations and climate
   across the northern tree line ecotone also seems more valuable preliminary exercise and one
   that should come before this paleo attempt at testing the model.

    In summary the considerable resources requested to develop very localised reconstructions
   of tree cover and basic climate for thousands of years is hard to justify when the likely
   impact of the results on future modelling studies is not likely to be great. Instead
   further sensitivity studies using a range of assumptions about localised albedo changes
   could be proposed using general circulation models or linking them to different vegetation
   models.
   (b) Fit to NERC priorities C
   (c) Risk Reward Medium for both Risk and Reward
   (d) Cost Effectiveness 3
   -------------------------------------------------------------------------------------------
   ----------------------------------------------------

   Professor Keith Briffa,
   Climatic Research Unit
   University of East Anglia
   Norwich, NR4 7TJ, U.K.

   Phone: +44-1603-593909
   Fax: +44-1603-507784
