NOAA Paleoclimatology Reconstructions Network ----------------------------------------------------------------------- World Data Center for Paleoclimatology, Boulder and NOAA Paleoclimatology Program ----------------------------------------------------------------------- NOTE: PLEASE CITE ORIGINAL REFERENCES WHEN USING THIS DATA!!!!! NAME OF DATA SET: NOAA Paleoclimatology Reconstructions Network RELEASE: 3/2009 (Version 1.0.0) UPDATE: 10/2009 (addition of recalibrations and wavelet plots) UPDATE: 12/2009 (updating of recalibration data files and associated R scripts) UPDATE: 1/2010 (addition of original article reference, for Version 1.0.1) UPDATE: 9/2010 (Version 2.0.0, addition of HadCRU instrumental temperature data and Mann et al. 2008/2009 proxy data series. See DESCRIPTION section for details) UPDATE: 12/2010 (Version 2.1.0, addition of NCEP/NCAR reanalysis. See DESCRIPTION section for details) CONTRIBUTORS: Eugene R. Wahl, David M. Anderson, Bruce A. Bauer, Rodney Buckner, Edward P. Gille, Wendy S. Gross, Michael Hartman, and Anju Shah, NOAA Paleoclimatology Program. IGBP PAGES/WDCA CONTRIBUTION SERIES NUMBER: 2009-021 WDC PALEO CONTRIBUTION SERIES CITATION: Wahl, E.R., et al. 2009. NOAA Paleoclimatology Reconstructions Network. IGBP PAGES/World Data Center for Paleoclimatology Data Contribution Series # 2009-021. NOAA/NCDC Paleoclimatology Program, Boulder CO, USA. ORIGINAL REFERENCE: Wahl, E.R., D.M. Anderson, B.A. Bauer, R. Buckner, E.P. Gille, W.S. Gross, M. Hartman, and A. Shah. 2010. An archive of high-resolution temperature reconstructions over the past 2+ millennia. Geochem. Geophys. Geosyst., 11, Q01001, doi:10.1029/2009GC002817. ABSTRACT (Wahl et al. 2010): High-resolution temperature reconstructions (typically annually to seasonally resolved) have played a key role in understanding paleoclimate immediately prior to the beginning of the instrumental record, especially when calibrated to form an extension of comparable instrumental data coverage (global, hemispheric, and regional). Such calibration allows the information in the instrumental record to be quantitatively extended backward in time in an objective way, enabling description of much longer term fluctuations in climate than possible with instrumental data alone, although with significantly increased uncertainties inherent in proxy-based reconstructions. This data brief describes a newly integrated archive of nearly all the high-resolution temperature reconstructions of the past 2+ millennia included in NOAA's National Climatic Data Center, from small-regional to global scale, which also have been recalibrated to a standard set of instrumental data. Examination of the spectral structure of the data is additionally provided. ADDITIONAL REFERENCES: Brohan, P., J.J. Kennedy, I. Harris, S.F.B. Tett, and P.D. Jones. 2006. Uncertainty Estimates in Regional and Global Observed Temperature Changes: A New Dataset from 1850. Journal of Geophysical Research 111:D12106. doi:10.1029/2005JD006548. Mann, M.E., Z. Zhang, M.K. Hughes, R.S. Bradley, S.K. Miller, S. Rutherford, R.S and F. Ni. 2008. Proxy-based reconstructions of hemispheric and global surface temperature variations over the past two millennia. Proceedings of the National Academy of Sciences 105(36): 13252-13257. Mann, M.E., Z. Zhang, S. Rutherford, R.S. Bradley, M.K. Hughes, D. Shindell, C. Ammann, G. Faluvegi and F. Ni. 2009. Global Signatures and Dynamical Origins of the Little Ice Age and Medieval Climate Anomaly. Science 326:1256-1260. GEOGRAPHIC REGION: Global PERIOD OF RECORD: Reconstructions - 2066 BC - 2006 AD Proxy Data - 8854BC - 2003 AD Instrumental Data - 1850 AD - 2009 AD DESCRIPTION: NOAA Paleoclimatology Paleoclimate Network (PCN) version 2.0.0, including 92 high-resolution temperature records over the past 2+ millennia, 1209 proxy data time series used by Mann et al. (2008, 2009), and instrumental data of the HadCRUT3 and HadCRUT3v combined global land and ocean gridded (5°x5°) surface temperature data sets, averaged to annual and seasonal time resolution by NOAA Paleoclimatology. The Reconstructions, Proxy, and Instrumental data sets are described in detail below. The RECONSTRUCTIONS include global, hemispheric, regional, and local reconstructions, generally with annual time-step resolution. The records come with many categories of metadata, including complete citations to original publications, seasonal period reconstructed (when applicable), anomaly period (if reconstructions are anomalies), latitude/longitude coverage, and URLs to the original NOAA Paleoclimate web pages from which the data were drawn. Each record is available as a separate ASCII file with fixed header and data formats, allowing machine reading of the data and time-step information. All the records together are also available in netCDF, ASCII, and Excel formats, including the complete metadata within the files themselves. The netCDF format version is provided in two files, the first with each record represented as a separate variable and the second as a time-by-study array in which each record is represented as a component of the study dimension. Sample scripts to open the netCDF versions in the R programming environment are also provided. The ASCII Format Description is as follows: 24 Header lines delimited by # character: # TITLE: (Taken from the reconstruction's NOAA-Paleo ftp page): additional information added to differentiate multiple time series from the same research/publication # URL1: Link to NOAA-Paleo page related to the reconstruction # URL2: Additional link to NOAA-Paleo page related to the reconstruction # VARIABLE_TYPE: Either "Temperature" or "Temperature Anomaly" # START_MONTH: (blank when not specified or the reconstruction is for an entire year) # END_MONTH: (blank when not specified or the reconstruction is for an entire year) # UNIT: Temperature or anomaly fundamental unit; generally degrees "C" or "K" # TIME_STEP: Time step of original data, generally annual # ANOMALY_PERIOD: Period in relation to which anomalies are calculated when VARIABLE_TYPE is "Temperature Anomaly" (blank otherwise) # START_YEAR: Year AD of beginning of the reconstruction as reported by authors ("NaN" is placed between this year and the beginning of reported numerical data in cases data were reported by authors in this way) # STOP_YEAR: Year AD of ending of the reconstruction as reported by authors # INVESTIGATORS: Researchers involved in making reconstruction # CITATION: Publication citation for data (Follows usage of journal "Ecology", in following format -- Author Year Title Publication Volume Pages, and DOI if available) # NORTH_BOUNDING_COORDINATE: All BOUNDING_COORDINATES are >= 0.000 (3 significant digits after the decimal) and then use N,S,E,W designator after each as appropriate # SOUTH_BOUNDING_COORDINATE: (For hemispheric and global reconstructions, full-hemisphere or global coordinates are given unless other BOUNDING_COORDINATES have been specified) # EAST_BOUNDING_COORDINATE: # WEST_BOUNDING_COORDINATE: # ELEVATION: In meters above sea level # DESCRIPTION_SUMMARY: Information about the reconstruction as defined by NOAA-Paleo # NOTES1: First notes line, e.g., concerning any alterations to the original presentation of the data that were made by NOAA-Paleo -- the following being a common change: "Year 0 through years xxxx BC have been changed to -1 AD thru -(xxxx+1) AD. This was done to conform to the AD/BC dating convention that there is no year 0, and to follow the convention that BC years are here reported as -AD years. # NOTES2: Second notes line # VERSION: March 2009, v 1.0.0 # RESERVED1: (Reserved for future use) # RESERVED2: (Reserved for future use) Data columns with headers: YearAD Temperature (can either be "Temperature" or "Anomaly") ("NaN" is used for missing values) RECONSTRUCTIONS DATA are contained in matrix data files pcn-v100.txt, pcn-v100.xls, pcn-v100.nc and pcn-array-v100.nc (Text, Excel, or netCDF format), and in 92 individual text files with filenaming convention of lead author and year of the original publication, example cook2000.txt. The individual text files are found at: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/reconstructions/pcn/textfiles/ Sample scripts to open the netCDF versions in the R programming environment are also provided at: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/reconstructions/pcn/ RECONSTRUCTION Recalibrations: These files contain the PCN reconstructions recalibrated to HadCRUT3v 5x5 degree temperature data. There are two seasonal splits available for the temperature recalibrations. One set splits the seasons based on the calendar year. This set contains annual, January-March(winter), April-June(spring), July-September(summer) and October-December(fall) periods. The other set contains the traditional meteorological seasonal splits starting with the December of the previous year. The meteorological seasons set contains annual, December-February(winter), March-May(spring), June-August(summer) and September-November(fall) periods. The annual data cover each calendar year and are the same in both sets of files. The recalibrations are provided in Text format at: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/reconstructions/pcn/recalibrations/ and ftp://ftp.ncdc.noaa.gov/pub/data/paleo/reconstructions/pcn/recalibrations-DJF/ (one file for each record containing all five recalibrations) with the naming convention of lead author, year and then '-recalibraion' or '-recalibrationDJF', e.g. ammann2007-recalibration. The data are also provided in netCDF format (one file for annual and for each of the eight different seasonal sets of recalibrations, with the records as separate variables). The netCDF files are named with 'ann' for annual, or the first letters of the three months that make up the season contained, e.g. PCN-v1-recalibration-JFM.nc contains January, February and March. Sample scripts to open the netCDF versions in the R programming environment are also provided and named to indicate which netCDF file they will open. The recalibrations were fitted by ordinary least-squares (OLS) regression using all years that had both reconstruction data and spatially-averaged HadCRU data appropriate to the spatial coverage of the reconstruction. One record (pollack1998) has too low time resolution to enable recalibration. Some records lacked comparable HadCRU data to use in recalibration for one or more of the seasonal time periods, and these cases are represented by NaN values for that season. The meaning of recalibration in cases where a reconstruction has a negative correlation with the corresponding HadCRU data is uncertain (potentially representing a spatial mismatch between original calibration data and the HadCRU data). For records where at least one of the annual or seasonal HadCRU averages and the original reconstruction exhibit this behavior, recalibrations are not provided (24 records for calendar year seasons and 17 records for meteorological seasons). Wavelets: The wavelet plots are provided at: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/reconstructions/pcn/wavelets/ They are all in pdf format and are named with the studyname followed by '-wavelet', e.g. cook2000-wavelet.pdf. In these plots the years axis indicates the number of years from the first year and not the year AD. The ge2003, overpeck1997, pollack1998, and yang2002 reconstructions do not have wavelet graphs because their time resolution is too low to allow the wavelet analysis. RECONSTRUCTION Data files: pcn-v100.txt - PCN reconstructions in a text file pcn-v100.xls - PCN reconstructions in an Excel file pcn-v100.nc - PCN reconstructions in netCDF format pcn-array-v100.nc - PCN reconstructions in netCDF arranged in one large array of time-by-study. PCN-v1-recalibration-ann.nc - Recalibrations over the HadCRU annual period PCN-v1-recalibration-JFM.nc - Recalibrations over the HadCRU Jan-March period PCN-v1-recalibration-AMJ.nc - Recalibrations over the HadCRU April-June period PCN-v1-recalibration-JAS.nc - Recalibrations over the HadCRU July-September period PCN-v1-recalibration-OND.nc - Recalibrations over the HadCRU October-December period PCN-v1-recalibration-DJF.nc - Recalibrations over the HadCRU December-February period PCN-v1-recalibration-MAM.nc - Recalibrations over the HadCRU March-May period PCN-v1-recalibration-JJA.nc - Recalibrations over the HadCRU June-August period PCN-v1-recalibration-SON.nc - Recalibrations over the HadCRU September-November period Scripts to read netCDFs into R: R-script-to-OPEN-pcn-v100-nc.txt R-script-to-OPEN-pcn-array-v100-nc.txt R-script-to-OPEN-PCN-v1-recalibration-ann-nc.txt R-script-to-OPEN-PCN-v1-recalibration-JFM-nc.txt R-script-to-OPEN-PCN-v1-recalibration-AMJ-nc.txt R-script-to-OPEN-PCN-v1-recalibration-JAS-nc.txt R-script-to-OPEN-PCN-v1-recalibration-OND-nc.txt R-script-to-OPEN-PCN-v1-recalibration-DJF-nc.txt R-script-to-OPEN-PCN-v1-recalibration-MAM-nc.txt R-script-to-OPEN-PCN-v1-recalibration-JJA-nc.txt R-script-to-OPEN-PCN-v1-recalibration-SON-nc.txt PROXY DATA were added to the PCN in September 2010, Version 2.0.0. The proxy data section of the PCN includes the 1209 time series used by Mann et al. in their recent reconstructions of hemispheric and global average temperatures (2008), and their gridded global reconstructions (2009). This accumulation is considered the most complete global proxy data set currently available for last 2+ millennium temperature reconstruction. The 1209 time series come in two versions: the original data as compiled by Mann et al., and each time series infilled as necessary by Mann et al. up to a common stopping year of 1998 using the regularized expectation maximization (RegEM) algorithm. Both the original and infilled versions of the data are provided in the three standard file formats, with overall metadata and header structure in the PCN format from the Mann et al. 2008 article. The metadata file containing information for all the individual time series as published by Mann et al. is also provided (ASCII and EXCEL formats). It should be noted that gridded temperature reconstructions in Europe (Luterbacher et al., 2004) are utilized as proxies in the Mann et al. reconstructions, and are thus included in the PCN proxy set. Paleoclimate Network of 1209 time series from Mann et al. (2008/2009). Both the original and infilled time series are provided. Available files include the Data Description and Format, Original Proxy data in Text, Excel 2007, or netCDF format, and Infilled Proxy data in Text, Excel 2007, or netCDF format. [Note that the Excel files require the 2007 version in order to access the large number of proxy data columns.] A sample script to open the netCDF versions in the R programming environment is also provided. To minimize the file size and column counts for the text and Excel format files, the sample depth counts for the proxy data were moved into separate data files. Available are the sample depth counts for the Original Proxy data in Text or Excel 2007 format, and sample depth counts for the Infilled Proxy data in Text or Excel 2007 format. The Mann et al. metadata describing the 1209 data series (1209proxynames.txt) are also available in both Text and Excel (1209proxynames.xls) formats. The following list provides codes for the basic data types in the metadata (cf. column D in the Excel format). Note that a "1" at the end of the code number indicates less than annually-resolved data. 2000 Luterbacher et al. Europe gridbox annual temperature reconstructions 3000 Tree composites/reconstructions 3001 Marine sediments 4000 Lake sediments 5000 Various composites or reconstructions or historical records 6000 Cave deposits 7000 Corals 7500 Schweingruber/Osborn latewood density tree ring gridbox series 8000 Ice cores 9000 ITRDB ring width and earlywood/latewood characteristic data For the 106 records with data codes 3000, 4000, 5000, 6000, 7000, and 8000 (and corresponding 3001, etc. codes) more specificity concerning the proxy type for a particular record can be obtained from the information in columns Z ("script-call-name"), B ("description"), C ("Data type"), and AA (original reference information). PROXY DATA file list: mann2008dataoriginal.txt mann2008dataoriginal.xlsx mann2008original.nc mann2008datainfilled.txt mann2008datainfilled.xlsx mann2008infilled.nc mann2008depthoriginal.txt mann2008depthoriginal.xlsx mann2008depthinfilled.txt mann2008depthinfilled.xlsx 1209proxynames.txt 1209proxynames.xls Script to read netCDFs into R: R-script-to-OPEN-PCN-v2-mann2008original-nc.txt INSTRUMENTAL DATA were added to the PCN in September 2010, Version 2.0.0. The instrumental data section of the PCN provides the HadCRUT3 and HadCRUT3v combined global land and ocean gridded (5°x5°) surface temperature data sets at annual and seasonal time resolution. [The "T3v " version of HadCRU contains variance corrections in relation to the "T3" version.] Both versions extend from 1850-2009. These data sets are often used for calibration and validation of regional and large-scale last 2+ millennium temperature reconstructions, once annualized or seasonalized from their native monthly time step. Along with the gridded data, the corresponding global, Northern Hemisphere, and Southern Hemisphere mean annual time series are provided in three standard file formats. NOAA Paleoclimatology created annual and seasonal averages of the HadCRUT3 and HadCRUT3v gridded monthly instrumental temperature data for 1850-2009 for the Paleoclimate Network. These averaged data are available here in Text, Excel, and netCDF format. Also available are the corresponding HadCRU global or hemispheric annual averages. The text and netCDF formats provide the gridded and mean series individually for both HadCRUT3 and HadCRUT3v. The Excel format provides the data in one file each for HadCRUT3 and HadCRUT3v. In parallel to the PCN Reconstruction data, there are two seasonal splits available for both HadCRUT3 and HadCRUT3v. One set splits the seasons based on the calendar year. This set contains January-March (winter), April-June (spring), July-September (summer) and October-December (fall) periods. The other set contains the traditional meteorological seasonal splits starting with the December of the previous year. The meteorological seasons set contains December-February (winter), March-May (spring), June-August (summer) and September-November (fall) periods. [For a given grid cell, annual and seasonal averages were computed when at least two-thirds of the time steps involved had data values (8 out of 12 months, annual; 2 out of 3 months, seasonal).] Sample scripts to open the netCDF versions in the R programming environment are also provided. Format for annual grid files are 36 rows x 72 columns starting at 87.5 N, 177.5 W, ending at 87.5 S, 177.5 E. INSTRUMENTAL DATA FILE List: PCN HadCRUT3 DATA FILES: hcrut3-allsets.xls hcrut3-AMJ.nc hcrut3-AMJ.txt hcrut3-ann.nc hcrut3-ann.txt hcrut3-DJF.nc hcrut3-DJF.txt hcrut3-GLann.nc hcrut3-GLann.txt hcrut3-JAS.nc hcrut3-JAS.txt hcrut3-JFM.nc hcrut3-JFM.txt hcrut3-JJA.nc hcrut3-JJA.txt hcrut3-MAM.nc hcrut3-MAM.txt hcrut3-NHann.nc hcrut3-NHann.txt hcrut3-OND.txt hcrut3-SHann.nc hcrut3-SHann.txt hcrut3-SON.nc hcrut3-SON.txt PCN HadCRUT3V DATA FILES: hcrut3V-allsets.xls hcrut3V-AMJ.nc hcrut3V-AMJ.txt hcrut3V-ann.nc hcrut3V-ann.txt hcrut3V-DJF.nc hcrut3V-DJF.txt hcrut3V-GLann.nc hcrut3V-GLann.txt hcrut3V-JAS.nc hcrut3V-JAS.txt hcrut3V-JFM.nc hcrut3V-JFM.txt hcrut3V-JJA.nc hcrut3V-JJA.txt hcrut3V-MAM.nc hcrut3V-MAM.txt hcrut3V-NHann.nc hcrut3V-NHann.txt hcrut3V-OND.nc hcrut3V-OND.txt hcrut3V-SHann.nc hcrut3V-SHann.txt hcrut3V-SON.nc hcrut3V-SON.txt Scripts to read netCDFs into R: R-script-to-OPEN-PCN-v2-hcrut3V-ann-nc.txt R-script-to-OPEN-PCN-v2-hcrut3V-NHann-nc.txt R-script-to-OPEN-PCN-v2-hcrut3V-JAS-nc.txt REANALYSIS DATA were added to the PCN in December 2010, Version 2.1.0. The Reanalysis data section of the PCN provides the NCEP/NCAR 2.5° x 2.5° gridded temperature reanalysis data sets at seasonal and annual time resolution, in netCDF and text format. Seasonal and annual means, or anomalies from the 1961 AD - 1990 AD base period, are provided, including both meteorological and calendar seasonal averages. The files are formatted alternatively with grids beginning at either the prime meridian or 180°W. Files with names ending in "-W" are arranged with data starting at 180°W and stepping sequentially through 177.5°E. Files lacking the "-W" in their names are arranged starting with data for 0°E and stepping sequentially through 2.5°W. In parallel to the other PCN data sets, there are two seasonal splits available for the NCEP/NCAR Reanalysis. One set splits the seasons based on the calendar year. This set contains January-March (winter), April-June (spring), July-September (summer) and October-December (fall) periods. The other set contains the traditional meteorological seasonal splits starting with the December of the previous year. The meteorological seasons set contains December-February (winter), March-May (spring), June-August (summer) and September-November (fall) periods. PCN NCEP/NCAR Reanalysis data files: ncepT-AMJ.nc ncepT-AMJ.txt ncepT-AMJ-W.nc ncepT-AMJ-W.txt ncepT-ann.nc ncepT-ann.txt ncepT-ann-W.nc ncepT-ann-W.txt ncepTanom-AMJ.nc ncepTanom-AMJ.txt ncepTanom-AMJ-W.nc ncepTanom-AMJ-W.txt ncepTanom-ann.nc ncepTanom-ann.txt ncepTanom-ann-W.nc ncepTanom-ann-W.txt ncepTanom-DJF.nc ncepTanom-DJF.txt ncepTanom-DJF-W.nc ncepTanom-DJF-W.txt ncepTanom-JAS.nc ncepTanom-JAS.txt ncepTanom-JAS-W.nc ncepTanom-JAS-W.txt ncepTanom-JFM.nc ncepTanom-JFM.txt ncepTanom-JFM-W.nc ncepTanom-JFM-W.txt ncepTanom-JJA.nc ncepTanom-JJA.txt ncepTanom-JJA-W.nc ncepTanom-JJA-W.txt ncepTanom-MAM.nc ncepTanom-MAM.txt ncepTanom-MAM-W.nc ncepTanom-MAM-W.txt ncepTanom-OND.nc ncepTanom-OND.txt ncepTanom-OND-W.nc ncepTanom-OND-W.txt ncepTanom-SON.nc ncepTanom-SON.txt ncepTanom-SON-W.nc ncepTanom-SON-W.txt ncepT-DJF.nc ncepT-DJF.txt ncepT-DJF-W.nc ncepT-DJF-W.txt ncepT-JAS.nc ncepT-JAS.txt ncepT-JAS-W.nc ncepT-JAS-W.txt ncepT-JFM.nc ncepT-JFM.txt ncepT-JFM-W.nc ncepT-JFM-W.txt ncepT-JJA.nc ncepT-JJA.txt ncepT-JJA-W.nc ncepT-JJA-W.txt ncepT-MAM.nc ncepT-MAM.txt ncepT-MAM-W.nc ncepT-MAM-W.txt ncepT-OND.nc ncepT-OND.txt ncepT-OND-W.nc ncepT-OND-W.txt ncepT-SON.nc ncepT-SON.txt ncepT-SON-W.nc ncepT-SON-W.txt