date: Fri Aug 18 15:34:50 2000
from: Phil Jones <p.jones@uea.ac.uk>
subject: Some ideas for captions
to: "Leslie Malone" <Malone_L@gateway.wmo.ch>, <BillSqoak@aol.com>,"Leslie Malone" <Malone_L@gateway.wmo.ch>


 Leslie, Bill,
   Here are a few ideas for captions.

  Boreholes on 1.14 in box. This could also say (needs to say). 
  
    Over the British Isles the borehole average from 26 sites agrees with
 observed annual temperatures for Central England, which extend back to
 1659. Instrumental temperatures are shown for each year and smoothed on the
 50-year timescale. The borehole average is the straighter line extending
 back to 1600. (It would seem best to remove the error bars around the
 corehole estimate).

   Millennial-long curves

  Comparisons are shown of three different reconstructions of temperatures
 for the millennium on the 50-year timescale. The red curve is based on a
 limited number of proxy records (trees, ice cores, corals etc, see spread
 2.??). The blue is based on a more extensive set (greater numbers since
 the 15th century), while the green is based on tree-ring density series
 from nearly 400 sites from high-latitude and high-elevations areas of the
 Northern Hemisphere.

  1.15 bottom left (left page)

   Annual average tempeartures for the hemispheres and the globe for 1856-1999.
 The smooth curve highlights variations on the decadal timescale. The zero
 value is equivalent to the average value for the 1961-90 period (i.e. for
 the globe 14 degrees C as discussed in the text).

 1.15 top right (left page)
  
    Seasonal cycle of hemispheric and global temperatures in an absolute
 sense for the 1961-90 period.

 1.15  top (right page)

    Patterns of change in annual temperature and precipitation explained
 by the linear trend over three periods this century (1901-44, 1945-76
 and 1977-99). The periods were chosen to cover the three different epochs 
 of global temperature change experienced during the 20th century. White
 areas depict areas where there are insufficient measurements to calculate
 trends. Spatial coverage is poorer earlier in the century.

  I've tried to make them simple, but still giving interested readers enough
 detail. We need to indicate base periods and the degree of smoothing.

   Captions need to be informative and maybe some terms can be used that
 would appear in the glossary. I'd be happy to hear views on my ideas.
 There could be some cross-referencing to other spreads.
 There is clearly a lot of work to be done on these as I'm only considering
 a few diagrams in a couple of spreads. 

   I found it hard to describe everything on the figures, a problem that
 will be very hard on some. Can Trevor remove some lines (eg the error
 bars on the borehole curve) ?
   
 Cheers
 Phil

