8. CONCLUDING REMARKS

The application of new technologies in information communication is taking the world by storm. While high speed data transmission has made the Web service of the Internet such a glamorous one, high speed has also benefited the email and FTP Services It now possible to FTP huge graphics files very quickly.

The WMO wants a fast, cheap, secure and manageable data communication system and is therefore thrilled with the prospects of applying the Internet technology to its GTS network. However there are several constraining factors elaborated above, some of which are recapitulated here as follows:

Evidentially, an NMC can not provide an accurate weather forecast based on erroneous data, just as correct late data is of little value to the forecaster.

Expertise is definitely a pre-requisite for anyone to benefit from high technology. So as operators in cyberspace trying to resolve the above constraints NMCs must strive to acquire the needed IT expertise. This needs to be strongly emphasised because without expertise NMCs may not make the most out of ongoing information communication revolution.

However, it is felt that many NMCs are in a position to make a start, probably on two parallel roads to improve meteorological data communications; namely the GTS road and the Internet-technology-based road as follows.

The GTS Road

The Internet Road

At any point in this migration process when resources permit, and when there is sufficient local expertise the Internet LAN and the GTS LAN can be interconnected after a firewall, like "Protocol Isolation with Server Replication " or the Microsoft "Catapult", is installed to protect the GTS LAN.

While cyberspace experts continue to seek solutions to Internet communication problems there is one of several major problems which only meteorological IT experts can solve namely the availability of cheap PC-based meteorological application programs. This problem is likely to slow down the GDPS computerisation process in RA I. RA I needs the equivalent of CLICOM in the area of AMSS, Real Time or Near Real Time Database Management, Operational Weather Forecasting, etc.

Within RA I, there are definitely NMCs that have excellent in-house developed applications that can greatly assist other NMCs if they are properly documented and made available to them. May be there is a need to make in-depth investigations of the availability of such software within the relatively more advanced RSMCs/NMCs in RA I. Maybe ACMAD should be tasked to do this, after all, one of ACMAD's slogans is "Application Development is an Obligation".

Another area that requires emphasising is the telecommunication links, especially those that connect NMCs to RTHs. It is not good enough for RSMCs/RTHs to generate/store lots of data/products if these can not reach the NMCs. There is a need to have at least medium speed links to facilitate graphics transmission/reception. Also there is a need to have "dual routes" to resources so that when one route fails the second one can be used as an alternative route.

The training seminar on the Use of New Technology for the Exchange and Processing of Meteorological Data and Products: RMTC Nairobi, Kenya, 8-26 May 2000, was excellent in content, implementation logistics notwithstanding. It was rather painful to hear, at the final plenary session, that the next implementation of a similar seminar in RA I will be for French Speaking NMCs two years hence. RA I need the communication expertise and the Nairobi seminar provided a great opportunity to acquire it.

For more then 3 years, DMC Nairobi had the ONLY DEC VAX 11/750 mini computer in Eastern Africa with no vendor support available within the region. But for all that time the system was kept running with no down time. DMC managed this because it had very well trained staff in that system hardware and software. On the other hand a Missir-Com/Vision system which has recently been installed at an NMC in RA I is experiencing problems and local staff are unable to solve even what may be simple problems. This is a rather unsatisfactory situation; NMCs require 80-90% local problem solving capacity.