[Section 7] [Introduction] [Documents] [Data] [Start]

8 The Berlin Data

8.1 The Directory Structure

8.2 Gridpoint Data Format

8.3 Zonal Mean Data

8.4 The QBO Data

8.5 Software


8 The Berlin Data

All Berlin data (1957 - 2001) are included on this CD.

Prior to 1973, the daily Berlin analysis charts were read out onto regular gridpoints manually. Therefore, all data prior to this year is available in a 10 x 10 horizontal resolution, and it covers the northern hemisphere northward of 10N only. At 80N, values are only sampled in a 20 longitudinal resolution.

Later, the interpolation onto gridpoints was done by a computer, allowing for a horizontal regular resolution of 5 x 5. In addition, the analysis was extented to the equator. To keep compatibility to the earlier data format, the 10 x 10 format is still maintained; however, it is now just a subset of the higher resolution data.

8.1 The Directory Structure

All data on this CD can be found in the directory data, which contains five directories 10x10, 5x5, longmean, zonalmean and qbo.

The diagram shows the directory tree beginning with data. As an example for all gridpoint data and the climatological data the structure of the directory height_30 is shown in detail. All other directories below data/10x10 and data/5x5 are organised in the same way, as far as daily are available.

Table 2 and Table 3 give an overview of all available gridpoint data.

data
    \
     |- 10x10
     |     \
     |      |- height_100
     |      |- height_50
     |      |- height_30
     |      |          \
     |      |           |- y57
     |      |           |- ...
     |      |           |- y95
     |      |           |     \
     |      |           |       - m9501.dat
     |      |           |      |- m9501
     |      |           |      |       \
     |      |           |      |        - d950101.dat
     |      |           |      |        - d950102.dat
     |      |           |      |        - ...........
     |      |           |      |        - d950131.dat
     |      |           |      |- .....
     |      |           |       - m9512.dat
     |      |           |      |- m9512
     |      |           |- y01
     |      |- height_10
     |      |- temp_100
     |      |- temp_50
     |      |- temp_30
     |      |- temp_10
     |- 5x5
     |     \
     |      |- height_100
     |      |- height_50
     |      |- height_30
     |      |- height_10
     |      |- temp_100
     |      |- temp_50
     |      |- temp_30
     |      |- temp_10
     |- longmean
     |       \
     |        |- 10x10
     |        |     \  
     |        |      |- height_100
     |        |      |- height_50
     |        |      |- height_30
     |        |      |           \
     |        |      |            |- m3001.dat
     |        |      |            |- m3701.dat
     |        |      |            |- .........
     |        |      |            |- m3012.dat
     |        |      |            |- m3712.dat
     |        |      |- temp_100
     |        |      |- temp_50
     |        |      |- temp_30
     |        |- 5x5
     |        |     \  
     |        |      |- height_100
     |        |      |- height_50
     |        |      |- height_30
     |        |      |- temp_100
     |        |      |- temp_50
     |        |      |- temp_30
     |- zonalmean
     |      \
     |       - bkm100h
     |       - bkm100t
     |       - bkm50h
     |       - bkm50t
     |       - bkm30h
     |       - bkm30t
     |       - bkm10h
     |       - bkm10t
     |- qbo
     |     \
     |      - qbo_53-01.dat
     |      - singapore_87-01.dat
  

8.2 Gridpoint Data Format

Please note the following points:

  • Geopotential heights are stored in gpm, temperatures in 0.1C.
  • Each monthly mean or daily field is stored in an individual file.
  • All fields cover the northern hemisphere only. The format of the data files are described in Section 8.2.1 and Section 8.2.2.

8.2.1 Format: 10 x 10

Data is available from 10N to 90N. In 80N, values are sampled every 20 degrees only! There is only one value for the North Pole (i.e., 90N).

The data is stored in the following order (F is either temperature or geopotential height):

F(10N,0E) , F(10N,10E) , ... , F(10N,350E) ,   <-- (36 values)
F(20N,0E) ,        ...       , F(20N,350E) ,   <-- (36 values)
F(30N,0E) ,        ...       , F(30N,350E) ,   <-- (36 values)
F(40N,0E) ,        ...       , F(40N,350E) ,   <-- (36 values)
F(50N,0E) ,        ...       , F(50N,350E) ,   <-- (36 values)
F(60N,0E) ,        ...       , F(60N,350E) ,   <-- (36 values)
F(70N,0E) ,        ...       , F(70N,350E) ,   <-- (36 values)
F(80N,0E) , F(80N,20E) , ... , F(80N,350E) ,   <-- (18 values) !
F(NP)                                          <-- ( 1 value ) !
                                                    ----------
                                        total     (271 values)
  

For all entries, a '(9I6)'-format is used; thus, each latitude circle consists of 4 rows in the data file, except from 80N (2 rows) and the pole (1 row, 1 value). The general format of each file is

line no. |  content
-------------------------
      1  |  header
      2  | \
      .  |  \
      .  |   > data lines
      .  |  /
     32  | /
  

The header line consists of 4 integer entries:

ilevid  iyear  imonth  iday
  

where the four(five) digit number ilevid has the structure

ilevid : (p)ppab  

         (p)pp     <-- pressure level (in hPa)
              a    <-- ID code: 1 for geopotential,
                                2 for temperature
               b   <-- ID code: 0 for daily values,
                                1 for monthly means.
  

iyear, imonth and iday denote the year, month, and day, respectively.

8.2.2 Format: 5 x 5

Data is available from 0N (equator) to 90N. There is only one value for the North Pole (i.e., 90N).

Data is stored in the following order (F is either temperature or geopotential height):

F( 0N,0E) , F( 0N,5E) , ... , F( 0N,355E) ,   <-- (72 values)
F( 5N,0E) , F( 5N,5E) , ... , F( 5N,355E) ,   <-- (72 values)
F(10N,0E) ,        ...      , F(10N,355E) ,   <-- (72 values)
F(15N,0E) ,        ...      , F(15N,355E) ,   <-- (72 values)
     .              .              .
     .              .              .
     .              .              .
F(70N,0E) ,        ...      , F(70N,355E) ,   <-- (72 values)
F(75N,0E) ,        ...      , F(75N,355E) ,   <-- (72 values)
F(80N,0E) ,        ...      , F(80N,355E) ,   <-- (72 values)
F(85N,0E) ,        ...      , F(85N,355E) ,   <-- (72 values)
F(NP)                                         <-- ( 1 value ) !
                                              ---------------
                                       total    (1297 values)
  

For all entries, a '(9I6)'-format is used; thus, each latitude circle consists of 8 rows in the data file, except from the pole (1 row, 1 value). The general format of each file is

line no. |  content
-------------------------
      1  |  header
      2  | \
      .  |  \
      .  |   > data lines
      .  |  /
    146  | /
  

The header line consists of 6 integer entries:

ilevel  iwhat  isample  iyear  imonth  iday
  

where

ilevel     <-- pressure level in hPa
iwhat      <-- 1 for geopotential height,
               2 for temperature
isample    <-- 0 for daily values,
               1 for monthly means
  

iyear, imonth and iday denote the year, month, and day, respectively.

8.2.3 Climatological Means

For the 100-, 50- and 30-hPa level climatological means for 30 years are available for temperature and geopotential height in both horizontal resolutions (cf.Section 3).

The format of the data section for these data is equal to the standard 10 x 10 (cf. Section 8.2.1) and 5 x 5 (cf. Section 8.2.2) format.

Depending on the horizontal resolution, the header is slightly different. For the 10 x 10 resolution the header line consists of 4 integer entries:

ilevid  iyear  imonth  iday
  

where the four(five) digit number ilevid has the structure

ilevid : (p)ppab  

         (p)pp     <-- pressure level (in hPa)
              a    <-- ID code: 1 for geopotential,
                                2 for temperature
               b   <-- ID code: 3 for longterm monthly mean
  

iyear, imonth and iday denote the start year, month and end year respectively.

For the 5 x 5 resolution the header line consists of 6 integer entries:

ilevel  iwhat  isample  isyear  imonth  ieyear
  

where

ilevel     <-- pressure level in hPa
iwhat      <-- 1 for geopotential height,
               2 for temperature
isample    <-- 3 for longterm monthly mean
  

isyear, imonth and ieyear denote the start year, month and end year respectively.

8.3 Zonal Mean Data

For the gridpoint data with 10 x 10 resolution all the monthly mean data are also available as zonal mean data.

The data for each variable on a certain level are stored in a single data file, starting with a header:

PPP:LLL, ELEMENT (1/2):I, ANF-JAHR:SYY, ANF-MONAT:SMM, END-JAHR:EYY, END-MONAT:EMM
  

where the following informations are coded:

LLL: Level in hPa
I: Variable (I=1 Geopotential Height in gpm, I=2 Temparature in 0.1C)
SYY: Start year of the data
SMM: Start month of the data
EYY: End year of the data
EMM: End month of the data

Within each data file the data are organized in columns, where the first column contains information about level, latitude, variable and year, and the remaining columns stand for each month of the year.

The first column is organised like this: PPPLIYY, where

PPP: Level in hPa
L: Latitude (e.g. L=1 stands for 10N)
I: Variable (I=1 Geopotential Height in gpm, I=2 Temparature in 0.1C)
YY: Year

Month with missing data are indicated with 0.

8.4 The QBO Data

The data in the file qbo_53-01.dat have the following format:

IIIII YYMM  70hPaN 50hPaN 40hPaN 30hPaN 20hPaN 15hPaN 10hPaN
  

where

IIIII=Station ID    (cf. Table 8)
YY=year             (53 - 01 => 1953 -  2001)
MM=month            (01 - 12)
N=1-9               if less than 10 daily values per month
N=0                 inter/extrapolated values
  

Table 8: Stations, used for the QBO data series.

Station IDStationLatitudeLongitude
91700 Canton Island 2 46' S171 43' W
43599 Gan/Maledives 0 41' S73 09' E
48694/698 Singapore 1 22' N103 55' E

The data in singapore_87-01.dat have the format of a time-height section for each single year.

In both data sets, the monthly mean zonal wind components are stored in 0.1 m/s; negative values are easterlies (from the east) and positive values are westerlies (qbo_53-01.dat).

8.5 Software

The subdirectory software contains some fortran source code to read the Berlin data analyses and the QBO data.

Currently, there are six fortran subroutines to read the Berlin data analyses: readd05.f, readd10.f, readm05.f, readm10.f, readc05.f and readc10.f to read daily, monthly and climatological data at 5 x 5 and 10 x 10 resolution, respectively. The sample Program readbd.f gives an example on how to use these subroutines and provides a possibility to convert the data to a format that can be used by GrADS (Grid Analysis and Display System). This program by Brian Doty is available for several operating systems, further information can be obtained at the GrADS Webpage.

The subroutine readbdzm.f can be used to read the zonal mean data. Again, the sample program readzm.f shows how it can be used.

To read the QBO data, two programs can be helpfull: readqbo.f to read the data from the file qbo_53-01.dat and read_singapore.f to read the data from the file singapore_87-01.dat.

Although we tried to check the functionality of the routines, we can not give any warranty for their usefulness or proper working on any kind of operating system. However, we would very much welcome any suggestions or bug reports. If you should have any problems with the data, do not hesitate to contact one of us, preferably via email.


[Section 7] [Documents] [Introduction] [Data] [Start]

Stratospheric Research Group
Last modified: Mon Sep 2 13:49:16 MEST 2002