 |
|
ATLAS TAPE/DISK DATASET DOCUMENTATION DEFINITIONS ATLAS7 describes the collection of data sets compiled in January 1993. All files except file 12 have been copied from ATLAS6. Night Marine Air Temperature anomalies (File 12) have been corrected. MOHSST4 describes Sea Surface Temperature (SST) data which are quality controlled with extreme values removed. MOHSST5 describes SST data which have been supplemented with values from the Comprehensive Ocean-Atmosphere Data Set (COADS) and additionally smoothed by spatial interpolation techniques. MOHMAT2N describes Night Marine Air Temperature (NMAT) data which are quality controlled with extreme values removed. COADS data are not included in this data set. DOCUMENTATION README : DOCUMENTATION FOR READING DATA AFTER CONVERTING TO ASCII Normals - SST data (MOHSST4) data) mohsst4_nor5dg_5180: 5 DEGREE AREA MONTHLY SST NORMALS FOR 1951-1980 mohsst4_nor1dg_5180: 1 DEGREE AREA MONTHLY SST NORMALS FOR 1951-1980 Normals - Air-Sea Temperature Difference Data (MOHSST4/MOHMAT2N data) asd_nor5dg_5180: 5 DEGREE AREA MONTHLY NORMALS OF AIR-SEA TEMPERATURE DIFFERENCE FOR 1951-80 Bucket data bws1991: "1991" BUCKET CORRECTIONS (WOODEN BUCKET - SHIP'S SPEED 4m/s) bwf1991: "1991" BUCKET CORRECTIONS (WOODEN BUCKET - SHIP'S SPEED 7m/s) bcs1991: "1991" BUCKET CORRECTIONS (CANVAS BUCKET - SHIP'S SPEED 4m/s) bcf1991: "1991" BUCKET CORRECTIONS (CANVAS BUCKET - SHIP'S SPEED 7m/s) Anomalies - SST data (MOHSST5 data) mohsst4_anm5dg_5691: 5 DEGREE AREA MONTHLY SST ANOMALIES, WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - ALL MONTHS 1856 TO 1991. sstanm_ts5691 : REGIONAL TIME SERIES OF SEASONAL SST ANOMALIES, WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - FROM 1856 TO 1991. sstanm_decseas: DECADAL AVERAGES OF SEASONAL SST ANOMALIES WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - FOR THE SEASONS JAN-MAR, APR-JUN, JUL-SEP, OCT-DEC AND THE DECADES 1861-70, 1871-80,.......1981-90. Anomalies - NMAT data (MOHMAT2N data) mohmat_anm5dg_5692: 5 DEGREE AREA MONTHLY NMAT ANOMALIES, WITH INSTRUMENT CORRECTIONS, W.R.T. 1951- 1980 - ALL MONTHS 1856 TO 1991 natanm_ts5692 : REGIONAL TIME SERIES OF SEASONAL NMAT ANOMALIES, WITH INSTRUMENT CORRECTIONS, W.R.T. 1951-1980 - FROM 1856 TO 1991. Statistical data (MOHSST4 data) sstcover_nrm5dg : TEMPORAL REPRESENTATIVENESS OF THE 5 DEGREE SST NORMALS IN FILE 2 I.E. THE NUMBER OF MONTHS IN EACH OF THE THREE DECADES 1951-1960 1961-1970 AND 1971-1980 WHICH CONTRIBUTE TO THE NORMALS IN A GIVEN 5 DEGREE AREA. sstnrm_range5dg : ANNUAL RANGE OF 5 DEGREE AREA SST NORMALS IN FILE 2 sstnrm_maxtim5dg: TIMES OF MAXIMA OF 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. sstnrm_mintim5dg: TIMES OF MINIMA OF 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. sst_standev5180 : STANDARD DEVIATIONS OF SST FOR THE PERIOD 1951 TO 1980 FOR EACH CALENDAR MONTH (5 DEGREE MOHSST REGION ONLY). sst_lagcorr5180 : ONE MONTH LAG CORRELATIONS OF SST ANOMALY FOR EACH CALENDAR MONTH FOR THE PERIOD 1951 TO 1980 (5 DEGREE MOHSST REGION ONLY). || || || NOTE: FILE 20 TO FILE 24 HAVE DIFFERENT FORMAT AND LAYOUTS || || FROM THE PREVIOUS FILES || || || Heat and Radiation data for the period 1949 to 1979 (From Mass. Inst. Tech.) heatflux_5dg: 5 DEGREE AREA LONG TERM MEAN OF NET HEAT FLUX. inrad_5dg : 5 DEGREE AREA LONG TERM MEAN OF INCOMING SOLAR RADIATION. outrad_5dg : 5 DEGREE AREA LONG TERM MEAN OF OUTGOING RADIATION. latent_5dg : 5 DEGREE AREA LONG TERM MEAN OF LATENT HEAT FLUX. sensible_5dg: 5 DEGREE AREA LONG TERM MEAN OF SENSIBLE HEAT FLUX. -------------------------------- DOCUMENTATION FOR USING THE TAPE FILE 1: readme7.txt (Dir: /CDROM/Document/Text) FILE 2: 5 DEGREE AREA SST AREA MONTHLY NORMALS FOR 1951-1980 (MOHSST4 Data) HDF File Name : A702sn5d.hdf ASCII Data File Name: A702sn5d.txt Long Data Set Name : MOHSST4_NOR5DG_5180 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NORMAL(72,36),MONTH DO 100 K=1,12 READ(10,1000)MONTH READ(10,2000)NORMAL 100 CONTINUE 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month for which the normals will be read into the array NORMAL. The normals are expressed in units of 0.01 degrees Celsius. Values for which reference has been made to the Alexander & Mobley (Mon.Weath. Rev.,104,143-148,1976) dataset in the derivation of the normal are flagged by the addition of 10000 i.e. the SST normal for such an area (I,J) = NORMAL(I,J)-10000(0.01 Celsius). Areas which have been designated as 'land' (i.e. all constituent 1 degree areas are land) or as 'ice' (i.e. more of the constituent 1 degree areas are ice than are sea) have the values -32768 and -1000 respectively. The layout of the normals array NORMAL is as follows: 90N ------------------------------- | | | | | | |<--------I=1,72|-------------->| 180W| | |180E | J=1,36 | | | | | | | ------------------------------- 90S i.e. for I=1, the 5 degree areas are centred on 177.5W, for I=2 on 172.5W & so on eastwards to I=72 on 177.5E, for J=1 the areas are centred on 87.5N, for J=2 on 82.5N & so on southwards to 87.5S. FILE 3: 1 DEGREE AREA MONTHLY SST NORMALS FOR 1951-1980 (MOHSST4 Data) HDF File Name : A703sn1d.hdf ASCII Data File Name: A703sn1d.txt Long Data Set Name : MOHSST4_NOR1DG_5180 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NORMAL(360,180),MONTH DO 100 K=1,12 READ(10,1000)MONTH DO 50 J=1,180 READ(10,2000)(NORMAL(I,J),I=1,180) READ(10,2000)(NORMAL(I,J),I=181,360) 50 CONTINUE 100 CONTINUE 1000 FORMAT(I6) 2000 FORMAT(180I6) where MONTH indicates the calendar month for which the normals will be read into the array ,NORMAL. The normals are expressed in units of 0.01 degrees Celsius. Areas which have been designated as 'land' have the value -32768 The layout of the array NORMAL is as follows: 90N -------------------------------------- | | | |<-----I=1,360-----|------------------>| 180W| | |180E | J=1,180 | | | | | | | -------------------------------------- 90S i.e. for I=1 the 1 degree areas are centred on 179.5W, for I=2 on 178.5W & so on eastwards to 179.5E; for J=1 the areas are centred on 89.5N, for J=2 on 88.5N & so on southwards to 89.5S. FILE 4: 5 DEGREE AREA MONTHLY NORMALS OF AIR-SEA TEMPERATURE DIFFERENCE FOR 1951-80 (MOHSST4/MOHMAT2N Data) HDF File Name : A704dn5d.hdf ASCII Data File Name: A704dn5d.txt Long Data Set Name : ASD_NOR5DG_5180 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ISDIFF(72,36),MONTH DO 100 K=1,12 READ(10,1000)MONTH READ(10,2000)ISDIFF 100 CONTINUE 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month for which the normals will be read into the array ISDIFF. The normals are expressed in units of 0.01 Celsius. Night marine air temperatures have been used to avoid the influence of solar heating of ships' decks; SSTs are MOHSST4 data and have been corrected to approximate night-time values by the subtraction of 0.1 deg Celsius. The missing data indicator is -32768 and the layout is as for File 2. FILE 5: "1991" BUCKET CORRECTIONS - WOODEN BUCKET SHIP'S SPEED 4m/s HDF File Name : A705bm04.hdf ASCII Data File Name: A705bxws.txt Long Data Set Name : BWS1991 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER IBUC(72,36) DO 20 IMON=1,12 READ(10,200)MONTH READ(10,300)IBUC 20 CONTINUE 200 FORMAT(I6) 300 FORMAT(72I6) where MONTH is the calendar month for which the corrections in IBUC are appropriate. The corrections are in 0.01 degrees Celsius. The missing data indicator is -32768 and the layout of the array is the same as for File 2. FILE 6: "1991" BUCKET CORRECTIONS - WOODEN BUCKET SHIP'S SPEED 7m/s HDF File Name : A705bm04.hdf ASCII Data File Name: A706bxwf.txt Long Data Set Name : BWF1991 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 7: "1991" BUCKET CORRECTIONS - CANVAS BUCKET SHIP'S SPEED 4m/s HDF File Name : A705bm04.hdf ASCII Data File Name: A707bxcs.txt Long Data Set Name : BCS1991 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 8: "1991" BUCKET CORRECTIONS - CANVAS BUCKET SHIP'S SPEED 7m/s HDF File Name : A705bm04.hdf ASCII Data File Name: A708bxcf.txt Long Data Set Name : BCF1991 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 9_*: 5 DEGREE AREA MONTHLY SST ANOMALIES WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2. ALL MONTHS 1856 to 1991. (MOHSST5 Data) EACH FILE REPRESENTS 10 YEARS OF DATA. (IE. A709sa85.hdf: begyear=1856 endyear=1865) HDF File Name : A709saXY.hdf (X=century Y=decade) ASCII Data File Name: A709saXY.txt (X=century Y=decade) Long Data Set Name : MOHSST5_ANM5DG_begyear_endyear The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO 100 IR=1,120 READ(10,1000)NMON,NYR READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. The anomalies are expressed in units of 0.01 Celsius. The missing data indicator is -32768. Instrumental corrections were grouped as follows. Proportion of wooden buckets (Pw) Pw = (1920 - Year) / 64 1854 to 1920 Pw = Zero 1921 to present Proportion of canvas buckets (Pc) Pc = One - Pw Proportion of slow (4 m/s) ships (Ps) Ps = One 1854 to 1870 Ps = (1940 - Year) / 70 1871 to 1940 Ps = Zero 1941 to present Proportion of fast (7 m/s) ships (Pf) Pf = One - Ps Corrections were applied in the following manner Corrections for wooden bucket & slow ship(File 5) * Pw * Ps + Corrections for wooden bucket & fast ship(File 6) * Pw * Pf + Corrections for canvas bucket & slow ship(File 7) * Pc * Ps + Corrections for canvas bucket & fast ship(File 8) * Pc * Pf No flags are stored in this data set. Initially all 5 degree areas were made up of at least 3 one-degree area pentad values. The remainder are set to to missing (-32768). These data have been smoothed after comparison with adjacent records. Each area can have up to eight spatially adjacent records and two temporally adjacent records. The layout of the array NDATA is as for the normals in File 2. FILE 9_13: 5 DEGREE AREA MONTHLY SST ANOMALIES WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2. ALL MONTHS 1986 to 1991. (MOHSST5 Data) HDF File Name : A709sa98.hdf ASCII Data File Name: A709sa98.txt Long Data Set Name : MOHSST5_ANM5DG_1986_1991 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO 100 IR=1,72 READ(10,1000)NMON,NYR READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. The anomalies are expressed in units of 0.01 Celsius. The missing data indicator is -32768. Instrumental corrections were grouped as follows. Proportion of wooden buckets (Pw) Pw = (1920 - Year) / 64 1854 to 1920 Pw = Zero 1921 to present Proportion of canvas buckets (Pc) Pc = One - Pw Proportion of slow (4 m/s) ships (Ps) Ps = One 1854 to 1870 Ps = (1940 - Year) / 70 1871 to 1940 Ps = Zero 1941 to present Proportion of fast (7 m/s) ships (Pf) Pf = One - Ps Corrections were applied in the following manner Corrections for wooden bucket & slow ship(File 5) * Pw * Ps + Corrections for wooden bucket & fast ship(File 6) * Pw * Pf + Corrections for canvas bucket & slow ship(File 7) * Pc * Ps + Corrections for canvas bucket & fast ship(File 8) * Pc * Pf No flags are stored in this data set. Initially all 5 degree areas were made up of at least 3 one-degree area pentad values. The remainder are set to to missing (-32768). These data have been smoothed after comparison with adjacent records. Each area can have up to eight spatially adjacent records and two temporally adjacent records. The layout of the array NDATA is as for the normals in File 2. FILE 10: REGIONAL TIME SERIES OF SEASONAL SST ANOMALIES, WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 ALL MONTHS 1856 TO 1991. (MOHSST5 Data) HDF File Name : A710sarg.hdf ASCII Data File Name: A710sarg.txt Long Data Set Name : SSTANM_TS5691 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ICHANG(20),XCHANG(540,20) DO 100 I=1,544 READ(10,1000)ICHANG DO 50 J=1,20 XCHANG(I,J)=ICHANG(J) 50 CONTINUE 100 CONTINUE 1000 FORMAT(20I6) where the indices I & J indicate the period and the area respectively, for which the anomaly held in array element XCHANG(I,J) ( or ICHANG(J)) is applicable: I PERIOD 1 Jan-Mar 1856 2 Apr-May 1856 . . . . 544 Oct-Dec 1991 J AREA (See also Figure 12 of the Atlas) 1 North Atlantic 2 South Atlantic 3 North Pacific 4 South Pacific 5 North Indian Ocean ------------------------------------------------------- 6 South Indian Ocean 7 Northern Tropics (0-20N) 8 Southern Tropics (0-20S) 9 Northern Hemisphere 10 Southern Hemisphere ------------------------------------------------------- 11 Atlantic Ocean 12 Pacific Ocean 13 Indian Ocean 14 Tropics (20N-20S) 15 Globe ------------------------------------------------------ 16 Tropical West Pacific 17 Tropical East Pacific 18 Tropical Pacific 19 Atlantic north of 35N 20 East Mid-latitude North Atlantic The anomalies are expressed in units of 0.01 degrees Celsius. Instrumental corrections have been applied as described in File 9. FILE 11: DECADAL AVERAGES OF SEASONAL SST ANOMALIES, WITH "1991" BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 FOR THE SEASONS JAN-MAR, APR-JUN, JUL-SEP, OCT-DEC, AND THE DECADES 1861-1870, 1871-1880, ...., 1981-1990. (MOHSST5 Data) HDF File Name : A711sadc.hdf ASCII Data File Name: A711sadc.txt Long Data Set Name : SSTANM_DECSEAS The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON1,NMON2,NYR1,NYR2,NDATA(72,36) DO 100 IR=1,52 READ(10,1000)NMON1,NMON2,NYR1,NYR2 READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(4I6) 2000 FORMAT(72I6) where NMON1 and NMON2 indicate the calendar months corresponding to the start and end of the season, and NYR1 and NYR2 are the years corresponding to the start and end of the decade for which the anomalies in the array NDATA are appropriate. The decadal seasonal means are expressed in units of 0.01 Celsius and, for a given 5 degree area , were calculated provided that seasonal means for 3 or more seasons (out of a possible 10) were available, so long as there was also at least one season available in each half of the decade. The intermediate seasonal means were calculated provided that data were available for 1 month or more (out of a possible 3). Instrumental corrections have been applied as described in File 9. The missing data indicator is -32768 and the layout of the data is as for normals in File 2. FILE 12_*: 5 DEGREE MONTHLY NMAT ANOMALIES WITH INSTRUMENT CORRECTIONS, W.R.T. 1951-1980 NORMALS, ALL MONTHS 1856 TO 1991 EACH FILES REPRESENT 10 YEARS OF DATA. (IE. A712aa85.hdf: begyear=1856 endyear=1865) HDF File Name : A712aaXY.hdf (X=Century, Y=Decade) ASCII Data File Name: A712aaXY.txt (X=Century, Y=Decade) Long Data Set Name : MOHMAT_ANM5DG_begyear_endyear The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO 100 IR=1,120 READ(10,1000)NMON,NYR READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. The anomalies are expressed in units of 0.01 Celsius. Instrumental corrections have been applied as described in the Global Ocean Surface Temperature Atlas Table 2 Section b(D). All 5 degree area monthly anomaly values have been created from at least 3 one-degree area anomalies. This version of NMAT data has replaced the ATLAS6 version, because data for the years 1982 to 1984 inclusive were incorrect. Also individual values in Febuary 1872 and January 1876 were found to be incorrect. We hope to produce a better version containing Quality Control flags showing results of range and variance tests within two years. The missing data indicator is -32768. The layout of the array NDATA is as for the SST normals in File 2. FILE 12_13: 5 DEGREE MONTHLY NMAT ANOMALIES WITH INSTRUMENT CORRECTIONS, W.R.T. 1951-1980 NORMALS, ALL MONTHS 1986 TO 1991 EACH FILES REPRESENT 10 YEARS OF DATA. (IE. A712aa98.hdf: begyear=1986 endyear=1991) HDF File Name : A712aa98.hdf ASCII Data File Name: A712aa98.txt Long Data Set Name : MOHMAT_ANM5DG_1986_1991 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO 100 IR=1,72 READ(10,1000)NMON,NYR READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. The anomalies are expressed in units of 0.01 Celsius. Instrumental corrections have been applied as described in the Global Ocean Surface Temperature Atlas Table 2 Section b(D). All 5 degree area monthly anomaly values have been created from at least 3 one-degree area anomalies. This version of NMAT data has replaced the ATLAS6 version, because data for the years 1982 to 1984 inclusive were incorrect. Also individual values in Febuary 1872 and January 1876 were found to be incorrect. We hope to produce a better version containing Quality Control flags showing results of range and variance tests within two years. The missing data indicator is -32768. The layout of the array NDATA is as for the SST normals in File 2. FILE 13: REGIONAL TIME SERIES OF SEASONAL NMAT ANOMALIES, WITH INSTRUMENT CORRECTIONS, W.R.T. 1951 TO 1980 NORMALS, FROM 1856 TO 1991 HDF File Name : A713aarg.hdf ASCII Data File Name: A713aarg.txt Long Data Set Name : NATANM_TS5691 The dataset may be read using the same Fortran routine as file 10. The anomalies are expressed in units of 0.01 Celsius. Instrumental corrections have been applied as described in the Global Ocean Surface Temperature Atlas Table 2 Section b(D). FILE 14: TEMPORAL REPRESENTATIVENESS OF THE 5 DEGREE NORMALS IN FILE 2 i.e. THE NUMBER OF MONTHS IN EACH OF THE THREE DECADES 1951-60,1961-70,1971-80 WHICH CONTRIBUTE TO THE NORMALS IN A GIVEN 5 DEGREE AREA. HDF File Name : A714sntm.hdf ASCII Data File Name: A714sntm.txt Long Data Set Name : SSTCOVER_NRM5DG The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER DATA(72,36),MONTH,IDEC DO 10 I=1,3 DO 100 K=1,4 READ(10,1000)MONTH,IDEC READ(10,2000)DATA 100 CONTINUE 10 CONTINUE 1000 FORMAT(2I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month (2,5,8,11 for Feb, May, Aug, Nov respectively) and IDEC indicates the decade (1,2,3 for 1951-60, 1961-70, 1971-80 respectively) for which the data will be read into the array DATA. The contributions have been counted only for those 5 degree areas for which the normal was derived purely from MOHSST4 data (such areas belong to what will hereafter be referred to as the 5 degree MOHSST4 region). The missing data indicator is -32768 and the layout of the array DATA is as for the normals on File 2. FILE 15: ANNUAL RANGE OF 5 DEGREE AREA SST NORMALS IN FILE 2 HDF File Name : A715sm04.hdf ASCII Data File Name: A715bxrn.txt Long Data Set Name : SSTNRM_RANGE5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NRANGE(72,36) READ(10,1000)NRANGE 1000 FORMAT(72I6) where array NRANGE contains the annual range of SST normal for each 5 degree area. The annual range is expressed in units of 0.01 degrees Celsius. Within the 5 degree MOHSST4 region, pentad data have been used to determine the annual range ,and monthly data elsewhere. The missing data indicator (i.e. land or ice) is -32768 and the layout of NRANGE is as for the array of 5 degree normals in File 2. See following two files for the times of the corresponding maxima and minima. FILE 16: TIMES OF MAXIMA OF 5 DEGREE AREA SST NORMALS: BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. HDF File Name : A715sm04.hdf ASCII Data File Name: A716snmx.txt Long Data Set Name : SSTNRM_MAXTIM5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MAXTIM(72,36) READ(10,1000)MAXTIM 1000 FORMAT(72I6) where MAXTIM contains the time of the maximum SST normal for each 5 degree area. The time of the maximum is given as a pentad number i.e. 1 to 73, or ( month number + 100) i.e. 101 to 112. (See description of annual range dataset on File 22.) If the maximum is maintained over a number of pentads/months the time of the maximum is given as the mid-point of the range.If the maximum value is repeated (i.e. in non- consecutive pentads/months ) the mid-point of the longest occurrence is given and such points are flagged with a negative sign. The missing data indicator (i.e. for land or ice) is -32768 and the layout of MAXTIM is as for the array of 5 degree normals in File 2. FILE 17: TIMES OF MINIMA OF 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. HDF File Name : A715sm04.hdf ASCII Data File Name: A717snmn.txt Long Data Set Name : SSTNRM_MINTIM5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MINTIM(72,36) READ(10,1000)MINTIM 1000 FORMAT(72I6) where MINTIM contains the time of the minimum SST normal for each 5 degree area. See the description of the dataset of times of maxima (File 16) for details. FILE 18: STANDARD DEVIATIONS OF SST FOR THE PERIOD 1951-80 FOR EACH CALENDAR MONTH (5 DEGREE MOHSST REGION ONLY) HDF File Name : A718snsd.hdf ASCII Data File Name: A718snsd.txt Long Data Set Name : SST_STANDEV5180 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MONTH,NDATA(72,36) DO 100 K=1,12 READ(10,1000)MONTH READ(10,2000)NDATA 100 CONTINUE 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month for which the standard deviations will be read into the array, NDATA. The standard deviations are expressed in units of 0.01 degrees Celsius and, for a given 5 degree area, were calculated provided that data were available for 16 months or more out of a possible 30. No instrumental corrections have been applied to the data. The missing data indicator is -32768. The layout of the data is as for the 5 degree normals in File 2. FILE 19: ONE MONTH LAG CORRELATIONS OF SST ANOMALY FOR EACH CALENDAR MONTH FOR THE PERIOD 1951-80 (5 DEGREE MOHSST REGION ONLY) HDF File Name : A719snlg.hdf ASCII Data File Name: A719snlg.txt Long Data Set Name : SST_LAGCORR5180 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NCORR(72,36),MONTH DO 100 K=1,12 READ(10,1000)MONTH READ(10,2000)NCORR 100 CONTINUE 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the first of the two successive calendar months for which the array NCORR of 1-month lag correlations is appropriate. The correlations, multiplied by a factor of 100, are given provided that data were available for 16 or more years in both of the successive calendar months in a 5 degree area. The missing data indicator is -32768 and the layout of NCORR is as for the array of 5 degree normals in File 2. FILE 20: 5 DEGREE AREA LONG TERM MEAN OF NET HEAT FLUX FOR 1949-1979 HDF File Name : A720rm05.hdf ASCII Data File Name: A720rxhx.txt Long Data Set Name : HEATFLUX_5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ARRAY(72,36),MONTH DO K=1,12 READ(10,1000)MONTH READ(10,2000)ARRAY ENDDO 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month and ARRAY is the LONGITUDE by LATITUDE array containing the long term means. Longitude dimension starts at 180W, increasing eastwards by a step of 5 degrees. Latitude dimension starts at 90N, increasing southwards by 5 degrees. The long term means are expressed in units of 0.1*watts/m**2. Missing data is denoted by the value -32768. FILE 21: 5 DEGREE AREA LONG TERM MEAN OF INCOMING SOLAR RADIATION FOR 1949-1979 HDF File Name : A720rm05.hdf ASCII Data File Name: A721rxir.txt Long Data Set Name : INRAD_5dg The read program and data layout are the same as FILE 20. FILE 22: 5 DEGREE AREA LONG TERM MEAN OF OUTGOING RADIATION FOR 1949-1979 HDF File Name : A720rm05.hdf ASCII Data File Name: A722rxor.txt Long Data Set Name : OUTRAD_5dg The read program and data layout are the same as FILE 20. FILE 23: 5 DEGREE AREA LONG TERM MEAN OF LATENT HEAT FOR 1949-1979 HDF File Name : A720rm05.hdf ASCII Data File Name: A723rxlt.txt Long Data Set Name : LATENT_5dg The read program and data layout are the same as FILE 20. FILE 24: 5 DEGREE AREA LONG TERM MEAN OF SENSIBLE HEAT FOR 1949-1979 HDF File Name : A720rm05.hdf ASCII Data File Name: A724rxsn.txt Long Data Set Name : SENSIBLE_5dg The read program and data layout are the same as FILE 20. *****************************************************************: END OF DOCUMENTATION *****************************************************************: ATLAS TAPE DATASET DOCUMENTATION DEFINITIONS ATLAS8 describes the collection of marine data sets compiled in January 1996. DOCUMENTATION README : DOCUMENTATION FOR USING THE TAPE (This listing) Normals - SST (GISST2.2) data gisst2_nor5dg_6190: 5 DEGREE AREA MONTHLY SST NORMALS FOR 1961-1990 gisst2_nor1dg_6190: 1 DEGREE AREA MONTHLY SST NORMALS FOR 1961-1990 Normals - Air-Sea Temperature Difference Data (MOHAST data) asd_nor5dg_6190: 5 DEGREE AREA MONTHLY NORMALS OF AIR-SEA TEMPERATURE DIFFERENCE FOR 1961-90 Bucket corrections (same as ATLAS7) bws1991: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS (WOODEN BUCKET - SHIP'S SPEED 4m/s) bwf1991: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS (WOODEN BUCKET - SHIP'S SPEED 7m/s) bcs1991: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS (CANVAS BUCKET - SHIP'S SPEED 4m/s) bcf1991: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS (CANVAS BUCKET - SHIP'S SPEED 7m/s) Anomalies - SST (MOHSST64 data) mohsst6_anm5dg_5695: 5 DEGREE AREA MONTHLY SST ANOMALIES, WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - ALL MONTHS 1856 TO 1995. sstanm_ts5695: REGIONAL TIME SERIES OF SEASONAL SST ANOMALIES, WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - FROM 1856 TO 1995. sstanm_decseas: DECADAL AVERAGES OF SEASONAL SST ANOMALIES WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 - FOR THE SEASONS JAN-MAR, APR-JUN, JUL-SEP, OCT-DEC AND THE DECADES 1861-70, 1866-75, 1871-80, 1876-85, ........... , 1986-95 Anomalies - NMAT data (MOHMATN41 data, "NIGHT" DEFINED AS SUNSET + 1 HOUR TO SUNRISE + 1 HOUR) mohmat_anm5dg_5695: 5 DEGREE AREA MONTHLY NMAT ANOMALIES, WITH CORRECTIONS AS IN PARKER ET AL "CLIMATE CHANGE" 1995, W.R.T. 1961- 1990 - ALL MONTHS 1856 TO 1995 natanm_ts5695 : REGIONAL TIME SERIES OF SEASONAL NMAT ANOMALIES, WITH CORRECTIONS AS IN PARKER ET AL "CLIMATE CHANGE" 1995, W.R.T. 1961-1990 - FROM 1856 TO 1995. Statistical data (MOHSST6 data) sstcover_nrm5dg : TEMPORAL REPRESENTATIVENESS OF THE 5 DEGREE SST NORMALS IN FILE 2 I.E. THE NUMBER OF MONTHS IN EACH OF THE THREE DECADES 1961-1970, 1971-1980 AND 1981-1990 WHICH CONTRIBUTE TO THE NORMALS IN A GIVEN 5 DEGREE AREA. sstnrm_range5dg : ANNUAL RANGE OF PENTAD 5 DEGREE AREA SST NORMALS UNDERLYING FILE 2 sstnrm_maxtim5dg: TIMES OF MAXIMA OF PENTAD 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. sstnrm_mintim5dg: TIMES OF MINIMA OF 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. sst_standev6190 : STANDARD DEVIATIONS OF SST FOR THE PERIOD 1961 TO 1990 FOR EACH CALENDAR MONTH (5 DEGREE MOHSST6 REGION ONLY). sst_lagcorr6190 : ONE MONTH LAG CORRELATIONS OF SST ANOMALY FOR EACH CALENDAR MONTH FOR THE PERIOD 1961 TO 1990 (5 DEGREE MOHSST6 REGION ONLY). Heat and Radiation data for the period 1949 to 1979 (From Mass. Inst. Tech.) heatflux_2dg: 5 DEGREE AREA LONG TERM MEAN OF NET HEAT FLUX. inrad_2dg : 5 DEGREE AREA LONG TERM MEAN OF INCOMING SOLAR RADIATION. outrad_2dg : 5 DEGREE AREA LONG TERM MEAN OF OUTGOING RADIATION. latent_2dg : 5 DEGREE AREA LONG TERM MEAN OF LATENT HEAT FLUX. sensible_2dg: 5 DEGREE AREA LONG TERM MEAN OF SENSIBLE HEAT FLUX. ------------------------------------------------------------ FILE 1: readme8.txt (Dir: /CDROM/Document/Text) FILE 2: 5 DEGREE AREA SST AREA MONTHLY NORMALS FOR 1961-1990 These normals are created from GISST2.2 data HDF File Name : A802sn5d.hdf ASCII Data File Name: A802sn5d.txt Long Data Set Name : GISST22_NOR5DG_6190 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NORMAL(72,36),MONTH DO K=1,12 READ(10,1000)MONTH,IYEAR READ(10,2000)NORMAL ENDDO 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month for which the normals will be read into the array NORMAL. The normals are expressed in units of 0.01 degrees Celsius. Areas which have been designated as 'land' (i.e. all constituent 1 degree areas are land) or as 'ice' (i.e. more of the constituent 1 degree areas are ice than are sea) have the values -32768 and -1000 respectively. These normals are documented by Parker et al in Hadley Centre Climate Research Technical note no. XXXX The layout of the normals array NORMAL is as follows: 90N ------------------------------- | | | | | | |<--------I=1,72|-------------->| 180W| | |180E | J=1,36 | | | | | | | ------------------------------- 90S i.e. for I=1, the 5 degree areas are centred on 177.5W, for I=2 on 172.5W & so on eastwards to I=72 on 177.5E, for J=1 the areas are centred on 87.5N, for J=2 on 82.5N & so on southwards to 87.5S. FILE 3: 1 DEGREE AREA MONTHLY SST NORMALS FOR 1961-1990 These normals are created from GISST2.2 data HDF File Name : A803sn1d.hdf ASCII Data File Name: A803sn1d.txt Long Data Set Name : GISST22_NOR1DG_6190 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NORMAL(360,180),MONTH DO K=1,12 READ(10,1000)MONTH DO J=1,180 READ(10,2000)(NORMAL(I,J),I=1,360) ENDDO ENDDO 1000 FORMAT(I6) 2000 FORMAT(360I6) where MONTH indicates the calendar month for which the normals will be read into the array ,NORMAL. The normals are expressed in units of 0.01 degrees Celsius. Areas which have been designated as 'land' have the value -32768 The layout of the array NORMAL is as follows: 90N -------------------------------------- | | | |<-----I=1,360-----|------------------>| 180W| | |180E | J=1,180 | | | | | | | -------------------------------------- 90S i.e. for I=1 the 1 degree areas are centred on 179.5W, for I=2 on 178.5W & so on eastwards to 179.5E; for J=1 the areas are centred on 89.5N, for J=2 on 88.5N & so on southwards to 89.5S. FILE 4: 5 DEGREE AREA MONTHLY NORMALS OF AIR-SEA TEMPERATURE DIFFERENCE FOR 1961-90 (MOHAST Data) HDF File Name : A804dn5d.hdf ASCII Data File Name: A804dn5d.txt Long Data Set Name : ASD_NOR5DG_6190 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ISDIFF(72,36),MONTH DO K=1,12 READ(10,1000)MONTH READ(10,2000)ISDIFF ENDDO 1000 FORMAT(I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month for which the normals will be read into the array ISDIFF. The normals are expressed in units of 0.01 Celsius. Obervations have been taken at night (1 hour after sunset to 1 hour after sunrise) to avoid the influence of solar heating of ships' decks; The missing data indicator is -32768 and the layout is as for File 2. FILE 5: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS - WOODEN BUCKET SHIP'S SPEED 4m/s HDF File Name : A805bm04.hdf ASCII Data File Name: A805bxws.txt Long Data Set Name : BWS1995 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER IBUC(72,36) DO IMON=1,12 READ(10,300)IBUC ENDDO 200 FORMAT(I6) 300 FORMAT(72I6) where MONTH is the calendar month for which the corrections in IBUC are appropriate. The corrections are in 0.01 degrees Celsius. The missing data indicator is -32768 and the layout of the array is the same as for File 2. FILE 6: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS - WOODEN BUCKET SHIP'S SPEED 7m/s HDF File Name : A805bm04.hdf ASCII Data File Name: A806bxwf.txt Long Data Set Name : BWF1995 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 7: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS - CANVAS BUCKET SHIP'S SPEED 4m/s HDF File Name : A805bm04.hdf ASCII Data File Name: A807bxcs.txt Long Data Set Name : BCS1995 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 8: FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS - CANVAS BUCKET SHIP'S SPEED 7m/s HDF File Name : A805bm04.hdf ASCII Data File Name: A808bxcf.txt Long Data Set Name : BCF1995 A description of the data set and Fortran routine statements to read the data can be found with the description of File 5. FILE 9_*: 5 DEGREE AREA MONTHLY SST ANOMALIES WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2. ALL MONTHS 1856 to 1995. (MOHSST64 Data) EACH FILES REPRESENT 10 YEARS OF DATA (IE. A809sa85.hdf: begyear=1856 endyear=1865) HDF File Name : A809saXY.hdf (X=century Y=decade) ASCII Data File Name: A809saXY.txt (X=century Y=decade) Long Data Set Name : MOHSST6_ANM5DG_begyear_endyear The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO IR=1,120 READ(10,1000)NMON,NYR READ(10,2000)NDATA ENDDO 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. The anomalies are expressed in units of 0.01 Celsius. The missing data indicator is -32768. Instrumental corrections were grouped as follows. Proportion of wooden buckets (Pw) Pw = (1920 - Year) / 64 1854 to 1920 Pw = Zero 1921 to present Proportion of canvas buckets (Pc) Pc = One - Pw Proportion of slow (4 m/s) ships (Ps) Ps = One 1854 to 1870 Ps = (1940 - Year) / 70 1871 to 1940 Ps = Zero 1941 to present Proportion of fast (7 m/s) ships (Pf) Pf = One - Ps Corrections were applied in the following manner Corrections for wooden bucket & slow ship(File 5) * Pw * Ps + Corrections for wooden bucket & fast ship(File 6) * Pw * Pf + Corrections for canvas bucket & slow ship(File 7) * Pc * Ps + Corrections for canvas bucket & fast ship(File 8) * Pc * Pf No flags are stored in this data set. Initially all 5 degree areas were made up of one-degree area pentad values. The remainder are set to to missing (-32768). These data have been smoothed after comparison with adjacent records. Each area can have up to eight spatially adjacent records and two temporally adjacent records. The layout of the array NDATA is as for the normals in File 2. FILE 10: REGIONAL TIME SERIES OF SEASONAL SST ANOMALIES, WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 ALL MONTHS 1856 TO 1995. (MOHSST64 Data) HDF File Name : A810sarg.hdf ASCII Data File Name: A810sarg.txt Long Data Set Name : SSTANM_TS5695 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ICHANG(20),XCHANG(560,20) DO I=1,560 READ(10,1000)ICHANG DO J=1,20 XCHANG(I,J)=ICHANG(J) ENDDO ENDDO 1000 FORMAT(20I6) where the indices I & J indicate the period and the area respectively, for which the anomaly held in array element XCHANG(I,J) ( or ICHANG(J)) is applicable: I PERIOD 1 Jan-Mar 1856 2 Apr-May 1856 . . . . 560 Oct-Dec 1995 J AREA (See also Figure 12 of the Atlas) 1 North Atlantic 2 South Atlantic 3 North Pacific 4 South Pacific 5 North Indian Ocean ------------------------------------------------------- 6 South Indian Ocean 7 Northern Tropics (0-20N) 8 Southern Tropics (0-20S) 9 Northern Hemisphere 10 Southern Hemisphere ------------------------------------------------------- 11 Atlantic Ocean 12 Pacific Ocean 13 Indian Ocean 14 Tropics (20N-20S) 15 Globe ------------------------------------------------------ 16 Tropical West Pacific 17 Tropical East Pacific 18 Tropical Pacific 19 Atlantic north of 35N 20 East Mid-latitude North Atlantic The anomalies are expressed in units of 0.01 degrees Celsius. Instrumental corrections have been applied as described in File 9. FILE 11: DECADAL AVERAGES OF SEASONAL SST ANOMALIES, WITH FOLLAND AND PARKER QJR METSOC 1995 BUCKET CORRECTIONS, W.R.T. THE NORMALS IN FILE 2 FOR THE SEASONS JAN-MAR, APR-JUN, JUL-SEP, OCT-DEC, AND THE DECADES 1861-70, 1866-75, 1871-80, 1876-85, ..... , 1986-95 (MOHSST64 Data) HDF File Name : A811sadc.hdf ASCII Data File Name: A811sadc.txt Long Data Set Name : SSTANM_DECSEAS The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON1,NMON2,NYR1,NYR2,NDATA(72,36) DO IR=1,104 READ(10,1000)NMON1,NMON2,NYR1,NYR2 READ(10,2000)NDATA ENDDO 1000 FORMAT(4I6) 2000 FORMAT(72I6) where NMON1 and NMON2 indicate the calendar months corresponding to the start and end of the season, and NYR1 and NYR2 are the years corresponding to the start and end of the decade for which the anomalies in the array NDATA are appropriate. The decadal seasonal means are expressed in units of 0.01 Celsius and, for a given 5 degree area , were calculated provided that seasonal means for 3 or more seasons (out of a possible 10) were available, so long as there was also at least one season available in each half of the decade. The intermediate seasonal means were calculated provided that data were available for 1 month or more (out of a possible 3). Instrumental corrections have been applied as described in File 9. The missing data indicator is -32768 and the layout of the data is as for normals in File 2. FILE 12_*: 5 DEGREE MONTHLY NMAT ANOMALIES WITH CORRECTIONS AS IN PARKER ET AL "CLIMATE CHANGE" 1995, W.R.T. 1961-1990 NORMALS, ALL MONTHS 1856 TO 1995 EACH FILES REPRESENT 10 YEARS OF DATA> (IE. A812aa85.hdf: begyear=1856 endyear=1865) HDF File Name : A812aaXY.hdf (X=Century, Y=Decade) ASCII Data File Name: A812aaXY.txt (X=Century, Y=Decade) Long Data Set Name : MOHMAT_ANM5DG_begyear_endyear The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NMON,NYR,NDATA(72,36) DO IR=1,120 READ(10,1000)NMON,NYR READ(10,2000)NDATA ENDDO 1000 FORMAT(2I6) 2000 FORMAT(72I6) where NMON indicates the calendar month and NYR the year for which the anomalies in array NDATA are appropriate. "Night" is defined as sunset + 1 hour to sunrise + 1 hour, the anomalies are expressed in units of 0.01 Celsius. Instrumental corrections have been applied as described in parker et al "climate change" 1995. All 5 degree area monthly anomaly values have been created from one-degree area pentad anomalies. The missing data indicator is -32768. The layout of the array NDATA is as for the SST normals in File 2. FILE 13: REGIONAL TIME SERIES OF SEASONAL NMAT ANOMALIES, WITH INSTRUMENT CORRECTIONS, W.R.T. 1961 TO 1990 NORMALS, FROM 1856 TO 1995 HDF File Name : A813aarg.hdf ASCII Data File Name: A813aarg.txt Long Data Set Name : NATANM_TS5695 "Night" is defined as sunset + 1 hour to sunrise + 1 hour. The dataset may be read using the same Fortran routine as File 10. The anomalies are expressed in units of 0.01 Celsius. Instrumental corrections have been applied as described in File 12. FILE 14: TEMPORAL REPRESENTATIVENESS OF THE 5 DEGREE NORMALS IN FILE 2 i.e. THE NUMBER OF MONTHS IN EACH OF THE THREE DECADES 1961-70,1971-80,1981-90 WHICH CONTRIBUTE TO THE NORMALS IN A GIVEN 5 DEGREE AREA. HDF File Name : A814sntm.hdf ASCII Data File Name: A814sntm.txt Long Data Set Name : SSTCOVER_NRM5DG The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER DATA(72,36),MONTH,IDEC DO I=1,3 DO K=1,4 READ(10,1000)MONTH,IDEC READ(10,2000)DATA ENDDO ENDDO 1000 FORMAT(2I6) 2000 FORMAT(72I6) where MONTH indicates the calendar month (2,5,8,11 for Feb, May, Aug, Nov respectively) and IDEC indicates the decade (1,2,3 for 1961-70, 1971-80, 1981-90 respectively) for which the data will be read into the array DATA. The contributions have been counted only for those 5 degree areas for which the normal was derived purely from MOHSST6 data (such areas belong to what will hereafter be referred to as the 5 degree MOHSST6 region). The missing data indicator is -32768 and the layout of the array DATA is as for the normals on File 2. FILE 15: ANNUAL RANGE OF PENTAD 5 DEGREE AREA SST NORMALS UNDERLYING FILE 2 HDF File Name : A815sm04.hdf ASCII Data File Name: A815bxrn.txt Long Data Set Name : SSTNRM_RANGE5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NRANGE(72,36) READ(10,1000)NRANGE 1000 FORMAT(72I6) where array NRANGE contains the annual range of SST normal for each 5 degree area. The annual range is calculated using pentad data and is expressed in units of 0.01 degrees Celsius. The missing data indicator is -32768 and the layout of NRANGE is as for the array of 5 degree normals in File 2. See following two files for the times of the corresponding maxima and minima. FILE 16: TIMES OF MAXIMA OF 5 DEGREE AREA SST NORMALS: BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. HDF File Name : A815sm04.hdf ASCII Data File Name: A816snmx.txt Long Data Set Name : SSTNRM_MAXTIM5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MAXTIM(72,36) READ(10,1000)MAXTIM 1000 FORMAT(72I6) where MAXTIM contains the time of the maximum SST normal for each 5 degree area. The time of the maximum is given as a pentad number i.e. 1 to 73. If the maximum is maintained over a number of pentads the time of the maximum is given as the mid-point of the range. If the maximum value is repeated (i.e. in non-consecutive pentads) the mid-point of the longest occurrence is given and such points are flagged with a negative sign. The missing data indicator (i.e. for land or ice) is -32768 and the layout of MAXTIM is as for the array of 5 degree normals in File 2. FILE 17: TIMES OF MINIMA OF 5 DEGREE AREA SST NORMALS; BASED ON PENTAD VALUES UNDERLYING THE MONTHLY NORMALS IN FILE 2. HDF File Name : A815sm04.hdf ASCII Data File Name: A817snmn.txt Long Data Set Name : SSTNRM_MINTIM5dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MINTIM(72,36) READ(10,1000)MINTIM 1000 FORMAT(72I6) Where MINTIM contains the time of the minimum SST normal for each 5 degree area. See the description of the dataset of times of maxima (File 16) for details. FILE 18: STANDARD DEVIATIONS OF SST FOR THE PERIOD 1961-90 FOR EACH CALENDAR MONTH (5 DEGREE MOHSST REGION ONLY) HDF File Name : A818snsd.hdf ASCII Data File Name: A818snsd.txt Long Data Set Name : SST_STANDEV6190 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER MONTH,NDATA(72,36) DO K=1,12 READ(10,1000)MONTH READ(10,2000)NDATA ENDDO 1000 FORMAT(I6) 2000 FORMAT(72I6) Where MONTH indicates the calendar month for which the standard deviations will be read into the array, NDATA. The standard deviations are expressed in units of 0.01 degrees Celsius and, for a given 5 degree area, were calculated provided that data were available for 16 months or more out of a possible 30. No instrumental corrections have been applied to the data. The missing data indicator is -32768. The layout of the data is as for the 5 degree normals in File 2. FILE 19: ONE MONTH LAG CORRELATIONS OF SST ANOMALY FOR EACH CALENDAR MONTH FOR THE PERIOD 1961-90 (5 DEGREE MOHSST REGION ONLY) HDF File Name : A819snlg.hdf ASCII Data File Name: A819snlg.txt Long Data Set Name : SST_LAGCORR6190 The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER NCORR(72,36),MONTH DO K=1,12 READ(10,1000)MONTH READ(10,2000)NCORR ENDDO 1000 FORMAT(I6) 2000 FORMAT(72I6) Where MONTH indicates the first of the two successive calendar months for which the array NCORR of 1-month lag correlations is appropriate. The correlations, multiplied by a factor of 100, are given provided that data were available for 16 or more years in both of the successive calendar months in a 5 degree area. The missing data indicator is -32768 and the layout of NCORR is as for the array of 5 degree normals in File 2. || || || NOTE: FILE 20 TO FILE 24 ARE ON A 2 DEGREE GRID || || || FILE 20: 2 DEGREE AREA LONG TERM MEAN OF NET HEAT FLUX FOR 1949-1990 HDF File Name : A820rm05.hdf ASCII Data File Name: A820rxhx.txt Long Data Set Name : HEATFLUX_2dg The ASCII Data File may be read using the following Fortran routine after it has been converted from the original HDF Data File: (For 'hdf2asci' conversion please go to the software directory) INTEGER ARRAY(180,90),MONTH DO K=1,13 READ(10,1000)MONTH READ(10,2000)ARRAY ENDDO 1000 FORMAT(I6) 2000 FORMAT(180I6) where MONTH indicates the calendar month and ARRAY is the LONGITUDE by LATITUDE array containing the long term means. Longitude dimension starts at 180W, increasing eastwards by a step of 2 degrees. Latitude dimension starts at 90N, increasing southwards by 2 degrees. The long term means are expressed in units of 0.1*watts/m**2. Missing data is denoted by the value -32768. FILE 21: 2 DEGREE AREA LONG TERM MEAN OF INCOMING SOLAR RADIATION FOR 1949-1990 HDF File Name : A820rm05.hdf ASCII Data File Name: A821rxir.txt Long Data Set Name : INRAD_2dg The read program and data layout are the same as FILE 20. FILE 22: 2 DEGREE AREA LONG TERM MEAN OF OUTGOING RADIATION FOR 1949-1990 HDF File Name : A820rm05.hdf ASCII Data File Name: A822rxor.txt Long Data Set Name : OUTRAD_2dg The read program and data layout are the same as FILE 20. FILE 23: 2 DEGREE AREA LONG TERM MEAN OF LATENT HEAT FOR 1949-1990 HDF File Name : A820rm05.hdf ASCII Data File Name: A823rxlt.txt Long Data Set Name : LATENT_2dg The read program and data layout are the same as FILE 20. FILE 24: 2 DEGREE AREA LONG TERM MEAN OF SENSIBLE HEAT FOR 1949-1990 HDF File Name : A820rm05.hdf ASCII Data File Name: A824rxsn.txt Long Data Set Name : SENSIBLE_2dg The read program and data layout are the same as FILE 20. *****************************************************************: END OF DOCUMENTATION *****************************************************************: