GISS Global Temperature Anomalies 1. TYPE OF DATA 1.1 Parameter/Measurement 1.2 Unit of Measurement 1.3 Data Source 1.4 Data Set Identification 2. SPATIAL CHARACTERISTICS 2.1 Spatial Coverage 2.2 Spatial Resolution 3. TEMPORAL CHARACTERISTICS 3.1 Temporal Coverage 3.2 Temporal Resolution 4. INSTRUMENT DESCRIPTION 4.1 Mission Objectives 4.2 Key Satellite Flight Parameters 4.3 Principles of Operation 4.4 Instrument Measurement Geometry 5. DATA PROCESSING SEQUENCE 5.1 Processing Steps and Data Sets 5.2 Derivation Techniques/Algorithms 5.3 Special Corrections/Adjustments 5.4 Processing Changes 6. QUALITY ASSESSMENT 6.1 Data Validation by Producer 6.2 Confidence Level/Accuracy Judgment 6.3 Usage Guidance 7. CONTACTS FOR DATA PRODUCTION INFORMATION 8. OUTPUT PRODUCTS AND AVAILABILITY 8.1 Tape Products 8.2 Film Products 8.3 Other Products 9. DATA ACCESS 9.1 Archive Identification 9.2 Procedures for Obtaining Data 9.3 NCDS Status/Plans 10. CONTACTS FOR ARCHIVE/DATA ACCESS INFORMATION 11. REFERENCES 11.1 Satellite/Instrument/Data Processing Documentation 11.2 Journal Articles and Study Reports 11.3 Archive/DBMS Usage Documentation 12. RELATED DATA SETS 13. SUMMARY/SAMPLE 14. NOTES ENDOFTOC 1. TYPE OF DATA 1.1 Parameter/Measurement. Surface air temperature deviations from the 1951-1980 average. 1.2 Unit of Measurement. Hundreths of degrees Centigrade (.01 deg C). 1.3 Data Source. The principal data sources for the surface air temperature station data are the World Weather Records (WWR), published by the Smithsonian Institution, and the WWR continuation, Monthly Climatic Data of the World (MCDW), collected under WMO (World Meteorological Organization) sponsorship, and published by NOAA's (National Oceanic and Atmospheric Administration) National Climatic Data Center (NCDC). 1.4 Data Set Identification. The GISS (Goddard Institute for Space Studies) monthly surface air temperature anomalies cover the period from January 1880 to December 1990. Anomalies are averaged over a range of spatial scales: a fine grid with 8000 equal-area cells, 80 equal-area regions, 8 latitude zones, 2 hemispheres, and the globe (see Item 2.2). 2. SPATIAL CHARACTERISTICS 2.1 Spatial Coverage. Near-Global coverage. 2.2 Spatial Resolution. The earth's surface is divided into 80 equal area "boxes", the full dimension of a box side being about 2500 km. For the locations of the boxes see Figure 2 in Hansen and Lebedeff (1987). Each of the 80 boxes is subdivided into an array of 10 by 10 equal-area "subboxes." The temperature anomaly for a subbox is defined using all stations located within 1200 km of that subbox's center. The 100 subbox values in a box are used to find the box average; the boxes are averaged to 8 latitude zones; and the zones are averaged to hemispheric and global anomalies. The table below shows the distribution among the 8 zones of the approximately 2000 continent and island meteorological stations used to compute the temperature anomalies. Surface Area Number of boxes No. of stations Latitude Zone Fraction in Latitude Band in Latitude Band ------------- ------------------ ---------------- ----------------- 64.2 - 90.0 N 0.05 4 88 44.4 - 64.2 N 0.10 8 461 23.6 - 44.4 N 0.15 12 495 0.0 - 23.6 N 0.20 16 437 0.0 - 23.6 S 0.20 16 260 23.6 - 44.4 S 0.15 12 179 44.4 - 64.2 S 0.10 8 29 64.2 - 90.0 S 0.05 4 18 The majority of ocean boxes contain at least several stations, but 5 of the 80 boxes contain no stations. Monthly anomalies are available for the globe, northern hemisphere, southern hemisphere, each of the zones listed above, each of the 80 equal-area boxes, and each of the 100 subboxes located within each of the larger boxes. Item 14 describes an accompanying program that converts monthly anomalies from the equal-area grid of 8000 subboxes to a uniform grid of 1 degree latitude by 1 degree longitude. 3. TEMPORAL CHARACTERISTICS 3.1 Temporal Coverage. Surface air station temperature anomalies are available from 1880 to 1990. Not all subboxes have data for the entire period. 3.2 Temporal Resolution. Monthly temperature anomalies. 4. INSTRUMENT DESCRIPTION N/A 5. DATA PROCESSING SEQUENCE 5.1 Processing Steps and Data Sets. The object of constructing this data set is to estimate the temperature change of large regions. Hansen and Lebedeff (1987) have sought to incorporate the information from all relevant available station records. However, in addition to limitations caused by incomplete spatial coverage, there are temporal inconsistencies that may be caused by changes in instrumentation, station location, observation time, or environmental factors. The spatial and temporal variability of the temperature anomalies has been screened to eliminate gross errors. 5.2 Derivation Techniques/Algorithms. Each of the 80 equal-area boxes is divided into a 10 by 10 array of subboxes. For each subbox, all stations within 1200 km are used to define the temperature change for that subbox. Each station's contribution to the temperature change for a given subbox is weighted according to the distance from the subbox center. All stations within 1200 km are ordered from the one with the longest temperature record to the one with the shortest record. The temperature changes for the first two stations are combined, then the third with the first two, and so on. Station records must have at least a 20-year overlap in order to be combined. The mean of both records is calculated for the period in common, then the entire second record is adjusted by the difference (bias). The mean of the resulting temperature records is the estimated temperature change as a function of time. Monthly temperature changes for a given subbox are obtained by applying this method to each of the twelve months. The temperature change for each of the 80 boxes is computed similarly by combining the temperature changes for the 100 boxes contained within the box, except that the subboxes are weighted equally and subboxes which have no station within 1200 km are excluded. The boxes in a latitude zone are likewise combined to find the temperature change for that zone. The contribution of a box to the zonal average is weighted by the proportion of the box's area which contains data. To compute the hemispheric temperature changes, first, the three highest latitude zones in a hemisphere (north of 23.6 N, south of 23.6) are combined to a single high latitude zone, applying the same method used to find the average for each of the eight zones. Then the temperature changes of the high latitude and the low latitude zones are combined, with each zone's contribution weighted by the zone's proportional area of the hemisphere. Temperature change for the globe is computed similarly from the two low latitude and two high latitude. 5.3 Special Corrections/Adjustments. N/A 5.4 Processing Changes. Work continues to find ways to improve this data set by adding station data, detecting and removing non-climatic influences on the station data, and refining the methods for evaluating temperature change for regions. 6. QUALITY ASSESSMENT 6.1 Data Validation by Producer. Station temperature deviations more than five standard deviations from the long term mean are excluded from the data unless the deviation is confirmed by data from neighboring stations. A quantitative estimate of the error due to imperfect spatial and temporal coverage of meteorological stations was obtained with the help of a 100-year run of a general circulation model (GCM). The GCM is described in Hansen et al., (1983). The goal was to find the error in the long-term surface air temperature change attributed to spatial gaps in station data. Of particular interest is the error in large-area results, such as the global surface air temperature change. 6.2 Confidence Level/Accuracy Judgment. The modeled and observed surface air temperature changes agreed reasonably well. At a 95 percent confidence level, the principal features in the global and hemispheric temperature changes are real, in the sense that they are not artifacts of poor spatial coverage of stations. The error caused by incomplete spatial coverage is estimated to be less that .1 degree C. 6.3 Usage Guidance. Spatial coverage is not complete over the entire globe. Station density is particularly sparse over mid- to high-latitude oceans in the Southern Hemisphere. Not all stations have data for the entire period 1880 through 1990. The procedure of Hansen and Lebedeff (1987) for defining temperature change is designed for obtaining the results for large regions, from the 1000 km scale to the global scale. Note that updated versions of this data set may contain some slightly different data values from a previous version. This occurs most frequently for data covering recent years, because of late reporting stations. Additionally, some data values may change slightly because the longer record permits better estimation of temperature changes. Therefore, when acquiring updates to this data set, it is recommended that the user not merely append a new year's worth of data to the earlier time series, but instead begin afresh with the full time series. 7. CONTACTS FOR DATA PRODUCTION INFORMATION Dr. Helene Wilson Center for the Study of Global Habitability Columbia University NASA/Goddard Institute for Space Studies 2880 Broadway New York, New York 10025 (212) 678-5640 BITNET: CEHWH@NASAGISS INTERNET: CEHWH@NASAGISS.GISS.NASA.GOV FAX: (212) 678-5552 8. OUTPUT PRODUCTS AND AVAILABILITY 8.1 Tape Products. Tapes can be made upon request. 8.2 Film Products. None. 8.3 Other products. 8.3.1 GISS Global Temperature Anomalies in Common Data Format (CDF) Medium/Specification: Magnetic Disk Format and Content: The data are stored in a special data-set-independent format, designated the Common Data Format (CDF). CDF was developed as a uniform method of storing and retrieving data on disk. The CDF contains data and descriptions about the data. A standard software package called the "CDF Library" allows a user within an NCDS session to create and access these data and descriptions. These CDF files may be accessed within the Data Applications subsystem of NCDS. Data Quantity/Rate: Data Set Name Volume ------------- ------------------ NCDS$DATA:GISS-TMP-DEV_MON-GLB 1 file, 22 Kb NCDS$DATA:GISS-TMP-DEV_MON-NH 1 file, 22 Kb NCDS$DATA:GISS-TMP-DEV_MON-SH 1 file, 22 Kb NCDS$DATA:GISS-TMP-DEV_MON-ZONE 1 file, 60 Kb NCDS$DATA:GISS-TMP-DEV_MON-BOXES 1 file, 441 Kb NCDS$DATA:GISS-TMP-DEV_MON-SUBBOXES 1 file, 42 Mb Status: Available Plans/Schedule: Additional data will be integrated periodically as they are received. 8.3.2 The Greenhouse Effect Detection Experiment (GEDEX) CD-ROM Medium/Specifications: CD-ROM in ISO 9660 standard Format and Content: The data on this disk will be stored in a special data-set-independent format, designated the Common Data Format (CDF). CDF was developed as a uniform method of storing and retrieving data on disk. The CDF contains data and descriptions about the data. A standard software package called the "CDF Library" will be included with this CD-ROM allowing a user to create and access these data and descriptions. The CD-ROM will consist of many different data sets related to the study of the Greenhouse Effect, including the GISS Temperature Deviations data sets, and the accompanying NCDS data set documentation. Data Quantity/Rate: 1 CD-ROM disk set. Status: Release date for this product is scheduled for early 1992. Plans/Schedule: Users interested in this product should contact NCDS staff for more details. 9. DATA ACCESS 9.1 Archive Identification. These data sets are available to users through the NASA Climate Data System (NCDS). For information about this data set or how to access NCDS, contact: Goddard DAAC User Support Office NASA/Goddard Space Flight Center Code 935 Greenbelt, Maryland 20771 (301) 286-3209 NSI DECnet: NSSDCA::NCDSUSO INTERNET: NCDSUSO@NSSDCA.GSFC.NASA.GOV 9.2 Procedures for Obtaining Data. The data may be obtained by online access through NCDS. Refer to Item 10 below for assistance in accessing the system. To obtain this data set in its native format, contact NASA Goddard Institute for Space Studies directly. (Refer to Item 7.) 9.3 NCDS Status/Plans. NCDS currently supports this data set in the Common Data Format (CDF) and will continue to update the data set as new data are received. (See Item 8.3.1 for Data Set Names). These files may be accessed by specifying the following data set names when entering the Data Applications subsystem of NCDS. Spatial Scale of Data Set Name Monthly Temperature Anomalies ------------- ----------------------------- NCDS$DATA:GISS-TMP-DEV_MON-GLB Global NCDS$DATA:GISS-TMP-DEV_MON-NH Northern Hemisphere NCDS$DATA:GISS-TMP-DEV_MON-SH Southern Hemisphere NCDS$DATA:GISS-TMP-DEV_MON-ZONE 8 Latitude Zones NCDS$DATA:GISS-TMP-DEV_MON-BOXES 80 equal area boxes NCDS$DATA:GISS-TMP-DEV_MON-SUBBOXES 8000 equal area subboxes 10. CONTACTS FOR ARCHIVE/DATA ACCESS INFORMATION For information on how to gain access to NCDS, contact the Goddard DAAC User Support Office: Goddard DAAC User Support Office NASA/Goddard Space Flight Center Code 935 Greenbelt, Maryland 20771 (301) 286-3209 NSI DECnet: NSSDCA::NCDSUSO INTERNET: NCDSUSO@NSSDCA.GSFC.NASA.GOV 11. REFERENCES 11.1 Satellite/Instrument/Data Processing Documentation. None. 11.2 Journal Articles and Study Reports a. Hansen, J., G. Russell, D. Rind, P. Stone, A. Lacis, S. Lebedeff, R. Ruedy and L. Travis, 1983. "Efficient Three-dimensional Global Models for Climate Studies: Models I and II." Mon. Wea. Rev., 111:609-662. b. Hansen, J. and S. Lebedeff, 1987. "Global Trends of Measured Surface Air Temperature." J. Geophys. Res., 92:13,345-13,372. c. Hansen, J. and S. Lebedeff, 1988. "Global Surface Air Temperatures: Update through 1987." Geophys. Res. Lett., 15:323-326. d. Jenne, R.L., 1975. "Data Sets for Meteorological Research." NCAR-TN/JA-111, National Center for Atmospheric Research, Boulder, CO. 192p. e. Spangler, W.M. and R.L. Jenne, 1990. "World Monthly Surface Station Climatology." Computer Data Tape Documentation, National Center for Atmospheric Research, Boulder, CO. 74p. 11.3 Archive/DBMS Usage Documentation a. Olsen, L.M., J.W. Closs, and F.E. Corprew, November 1991. "NASA's Climate Data System Primer: Version 4.0" EOS DAAC, NASA Goddard Space Flight Center, Greenbelt, Maryland. 12. RELATED DATA SETS Several related data sets exist online through NCDS. 1) The Climate Research Unit's Global Temperature Deviations data set contains surface air and marine air temperature anomalies derived from the World Weather Records (WWR). These data are available online through NCDS in Common Data Format (CDF). To access this CDF, specify the following as input within the Data Applications Subsystem in NCDS. NCDS$DATA:CLIM-RSRCH-U_TMP-DEV 2) The Angell Global Temperature Deviations data set contains zonal temperature, and temperature deviations derived from radiosonde measurements compiled in Monthly Climatic Data for the World, published by the National Climatic Data Center. These data are available online through NCDS in Common Data Format (CDF). To access this CDF, specify the following as input within the Data Applications Subsystem in NCDS. NCDS$DATA:ANGELL_TMP_DEV 3) Brightness Temperature Anomalies obtained from the Microwave Sounding Unit aboard the TIROS-N series of polar orbiting satellites (TIROS-N, NOAA-6, NOAA-7, NOAA-9, NOAA-10, NOAA-11, and NOAA-12). These data are available online through NCDS in Common Data Format (CDF). To access these CDFs, specify the following as input within the Data Applications subsystem of NCDS. NCDS$DATA:MSUTMP_MONTHLY NCDS$DATA:MSUTMP_CLIMATOLOGY 4) The Vinnikov, Groisman, and Lugina zonal and station temperature data sets contain monthly, seasonal, and annual mean zonal temperature deviations and monthly mean station temperature deviations. These data are available online through NCDS in Common Data Format (CDF). To access these CDFs, specify the following as input within the Data Applications subsystem of NCDS. NCDS$DATA:VGL-TMP-DEV_MONTHLY NCDS$DATA:VGL-TMP-DEV_SEASONAL NCDS$DATA:VGL-TMP-DEV_ANNUAL NCDS$DATA:VGL-TMP-MON_NORTH-STATN NCDS$DATA:VGL-TMP-MON_SOUTH-STATN 13. SUMMARY/SAMPLE A sample plot is available for this data set. It is accessible through the "Graph Data" option within the Data Applications Subsystem of NCDS. To view this plot, choose "Display Demonstration Plot of a Data Set" and specify the plot file NCDS$DEMO:GISS-TMP-DEV as input. 14. NOTES Included on the GEDEX CD-ROM is a program to replicate a month's 8000 subbox temperature anomalies to a uniform grid of 1 degree latitude by 1 degree longitude. The user must first subset the data using the CDFList utility available on the CD-ROM. (CDFList creates ASCII subsets of the data.) The program to create a uniform grid of 1 x 1 degree resolution reads the subbox data from the ASCII subset. The subboxes are organized in a sequence of month-by-month "maps." This program is not recommended for use to replicate the subbox values directly to a grid coarser than 1 by 1 degree, since some of the subbox values may be omitted with the coarser grid. A two-step procedure--from the subboxes to a 1 x 1 degree grid, and then interpolation to a coarser grid--is recommended. To access this program, specify... ...:GISS-TMP-DEV_MON-SUBBOXES-TO1X1GRID *92/03/13