![](bg2.png)
forecast fields produced with numerical models.
Moreover the progressive replacement of the
manned observing stations with automatic stations
has made the binary production of observation re-
ports more convenient and effective.
The continuous evolution of the WMO codes
and the heterogeneous nature of the observing sys-
tem have resulted in a very complex set of codes
which makes very difficult for a user to access the
information requested in a consistent way across the
various parts of the coding system. To improve
consistency across the codes the Table Driven Code
Forms (TDCF) (WMO, 2003) have been developed
to replace the traditional alphanumeric codes (TAC).
2 TAC VS. TDCF
There are 76 different traditional alphanumeric
codes (TAC) (WMO, 2012a). An example of TAC
message is the Aerodrome routine meteorological
report (METAR) (WMO, 2012a) which is the mes-
sage of observation produced every hour by the
airport observation station and used by the air traffic
control to inform the pilots on the weather condi-
tions at the airport. The code form is quite complex,
but is made to be interpreted by a trained operator.
Here follows an example:
METAR EDDF 120550Z 03015KT 1400
R07R/P2000N R07C/P2000N
R07L/1900U SN DRSN BR VV///
M04/M04 Q1000 R07L/11//90
R07C/15//90 R07R/15//90 BECMG
4000 NSW=
The code starts with the word METAR and ends
with the sing “=”. A set of numbers and letters are
divided in groups which are recognized for their
position and for their alphanumeric pattern. The
element EDDF is the four character code for the
airport, meaning that the observation refers to
Frankfurt airport. The second element 120550Z says
that the observation is for the 12
th
of the month at
5:50 UTC (the month and the year are not expressed
in the message). The third element 03015KT means
that the wind is coming from 30 degrees with a
speed of 15 Knots. It is clear that decoding of each
of the elements does not follow any general rule
except the fact that information is split in groups
with different meaning.
Another example of TAC is the SYNOP (Report
of surface observation from a fixed land station). An
example of this kind of message is the following:
AAXX 13094 03002 45462 /0514
10097 20073 30238 40256 58011
90850 333 88/11=
Where AAXX is the start of the report of this kind
and “=” is the end. Only numbers and “/” are al-
lowed in the body of the message, which are
grouped in groups of 5 with exception of the groups
marking new sections like the group 333 which
denotes the start of section 3. The first element
13094 means that the observation is valid for the
13
th
of the month (no explicit indication of month
and year) at 9 am and the number 4 means that the
wind speed is observed with an anemometer and is
reported in knots. Decoding the elements of the
message is out of our scope we only want to point
out that each group has a different meaning and each
number within the groups has quite complicated
decoding rules which are different group by group
and code figure by code figure.
A coding system like the one used in the TAC
without general rules, in which each element has
different decoding rules is difficult to extend, to
maintain and makes impossible the task of produc-
ing a general decoder. To overcome these limita-
tions of the alphanumeric codes it was decided to
produce a new system based on a unique set of ta-
bles providing a list of elements reusable in different
contexts. A unique set of rules to decode a message
were also provided in a form that is possible to im-
plement decoding software to access information
from the message. The fundamental concept of these
new codes called Table Driven Code Forms (TDCF)
was the separation between the coding rules and the
elements used in the code forms. The rules are ge-
neric and apply to all the different types of messag-
es, while the tables of elements and sequences are
provided as external support to the decoding soft-
ware in the form of a palette of elements to be re-
used in different contexts and a set of sequences of
elements with a special meaning to be used in the
definition of a single message. With the TDCF ap-
proach is possible to produce a decoder implement-
ing the decoding rules, which can take as input the
Tables and decode the message, allowing a big flex-
ibility and extensibility connected with the fact that
new messages can be defined through new elements
and sequences without changing the decoding soft-
ware.
WMO has developed two different TDCF named
BUFR (Binary Universal Form for the Representa-
tion of meteorological data) and GRIB (General
Regularly-distributed Information in Binary form).
To limit the scope of this paper we will consider
SENSORNETS2014-InternationalConferenceonSensorNetworks
410