Maiana Coral Oxygen Isotope Data: Readme file
---------------------------------------------------------------------
                NOAA Paleoclimatology Program
                               and
         World Data Center for Paleoclimatology, Boulder
---------------------------------------------------------------------
NOTE: PLEASE CITE ORIGINAL REFERENCE WHEN USING THIS DATA!!!!!

NAME OF DATA SET: Maiana Coral Oxygen Isotope Data
LAST UPDATE: 10/2000 (Original Receipt by WDC Paleo)

CONTRIBUTORS: Frank Urban, University of Colorado,
Julia Cole and Jonathan Overpeck, University of Arizona
	
IGBP PAGES/WDCA CONTRIBUTION SERIES NUMBER: 2000-063

SUGGESTED DATA CITATION: Urban, F. et al., 2000,
Maiana Coral Oxygen Isotope Data, IGBP PAGES/
World Data Center for Paleoclimatology 
Data Contribution Series #2000-063. 
NOAA/NGDC Paleoclimatology Program, Boulder CO, USA. 

ORIGINAL REFERENCE: Urban, F.E., J. E. Cole, and J.T. Overpeck, 2000, 
Influence of mean climate change on climate variability from a 
155-year tropical Pacific coral record. Nature 407, 989-993.

GEOGRAPHIC REGION: subtropical western Pacific
PERIOD OF RECORD: 1840-1995 AD

FUNDING SOURCES: National Science Foundation and 
National Oceanic and Atmospheric Administration, USA.

LIST OF FILES: Readme_maiana.txt (this file),
Maiana_18o.txt, Maiana_correlations.txt.
 

DESCRIPTION: 
Maiana bimonthly oxygen isotopic composition, 1840-1995

Notes on the data:
File includes columns for Year AD 
(bimonthly resolution = dec/jan, feb/mar…) and coral d18O (‰).

The coral used in this study grew (and still does grow) at Maiana Atoll, 
Republic of Kiribati (1°N, 173°E). The colony lies at a depth of ~6m (low tide) 
and is about 4m in height. Annual growth rate is roughly 11mm/yr. 

Methods:
Sampling and isotopic analysis: Cores (8.5 cm diameter) drilled from living coral 
colonies were slabbed and X-rayed to identify annual density bands and optimal 
sampling transects. Samples were drilled along the growth axis continuously at 
1mm resolution. Stable isotopic analysis (d18O and d13C) of powdered samples 
was performed on a Micromass Optima with Isocarb automated preparation system 
at the University of Colorado's Institute for Arctic and Alpine Research. 
Analytical precision is =±0.08 d18O and =±0.05 d13C. The Maiana record is 
continuous except for a gap from 1901-1905, which we bridged through 
straight forward comparison and splicing of data from a coral record from 
nearby Tarawa Atoll.

Age model: Seasonally varying d13C values and ephemeral annual density bands 
provide the basis for annual age assignment. Ages were adjusted subseasonally 
by comparison to major El Niño and La Niña anomalies in a regional rainfall 
index (available at http://tao.atmos. washington.edu/datasets/eqpacislandrainindex/); 
this procedure allows us to align major peaks and troughs that occur within the 
same calendar year with adjustments that are generally less than 3 months. 

Because corals grow at variable rates throughout the year, these small chronological 
adjustments are reasonable and customary (and do not affect estimates of multi-year 
variability). Near the base of the core, an age adjustment of 1 year was needed, 
suggesting an age uncertainty on the full record of ±1 year. The Maiana coral grew 
at a rate of 7-12 mm per year; we interpolated records to a common time step of 
bimonthly resolution for further analysis. 

Replication: The record from core MAI95-2-3 is well replicated by additional 
records of 65-70 years from Maiana and Aranuka, ~75 km southeast of Maiana, 
(JEC, FEU, and JTO, manuscript in preparation) and by a published 97-year 
record from Tarawa Atoll, ~30 km north of Maiana. Correlation coefficients 
among coral and instrumental data over the common period 1950-89 are consistently 
high (Maiana_correlations.txt), adding confidence to single-core reconstruction 
of regional variability in western Kiribati. We focus on the MAI95-2-3 record 
for further analysis because it extends the record of ENSO into pre-instrumental 
periods and avoids nonstationarity that may be introduced through inclusion of 
records that differ in length. The 4-coral average correlates slightly better 
with instrumental data than most individual coral records, but only Mai 2-3 
extends the instrumental record of ENSO. Linear correlation coefficients 
demonstrate a strong agreement between the Maiana record and ENSO variability 
(Table 1), and decadal averages of the Maiana record correlate significantly 
over the past century (>95% confidence) with those from the Tarawa coral 
record and regional rainfall indices, substantiating the reality of these 
more subtle low-frequency signals (JEC, FEU, and JTO, manuscript in preparation).