ABC of meteorological data encoding
Meteorological data around the globe are being exchanged among meteorological centres under the framework of the World Meteorological Organization (WMO) in a timely, reliable and cost-effective manner. They are of vital importance to the forecasting and warning operations of the Hong Kong Observatory. Currently, most meteorological observations are being produced and transmitted in special formats known as Traditional Alphanumeric Codes (TAC) and Table-Driven Code Forms (TDCF) as promulgated by WMO. Both TAC and TDCF are ways of representing meteorological data intended for international exchange in real-time.
Owing to limited bandwidth of telecommunication lines, meteorological data need to be encoded to reduce size. TAC is a combination of human readable alphanumeric characters developed many decades ago for transmission on low-bandwidth telecommunication lines, such as Telex. In order to reduce bandwidth consumption, abbreviated coding and translation by code rather than the exact value are two key features of TAC. Loss of precision in the data is therefore inevitable when using TAC. For example, surface observations encoded in TAC can only hold the visibility information in steps of 1 km for values above 5 km (see Figure 1).
Figure 1 Decoded TAC Information
Binary Universal Form for the Representation of meteorological data (BUFR) is a kind of TDCF developed by WMO as an intended replacement for TAC to encode meteorological observations in the late 80’s when the use of computers for meteorological data processing became more commonplace. BUFR is binary that makes it compact, flexible and expandable, and hence excellent for transmitting and storing huge amount of meteorological information. Nowadays, new types of meteorological information with higher resolution and measurement accuracy are emerging and exchanged among meteorological centres. The traditional TAC encoding method becomes incapable of accommodating the new data to meet the expansion and growing needs. BUFR, on the other hand, provides effective support for the representation of new data types, metadata, higher temporal or spatial resolution data and higher precision data such as meteorological satellite data, aircraft observations and also tropical cyclone information (Figure 2).
Figure 2 Decoded BUFR Information
BUFR is now an important component in meteorological information exchange. The benefits of using BUFR are becoming more well-known and recognized by the meteorological centres. The Hong Kong Observatory started to produce daily observations in BUFR format in 2010 and all essential observation reports are being disseminated to other meteorological centres in both TAC and BUFR formats.