Key to sustainable networks Manpower and skills more

Key to sustainable networks: Manpower and skills more then equipment J. Havskov and L. Ottemöller

Introduction • University of Bergen has been involved in many network projects, training workshops etc, mostly in developing countries • Our main observation is that too much emphasis has been put on equipment and too little on human resources and operational costs

Questions to be asked • Who wants a network: Local politicians, local scientists or external organizations • Who will operate, are the local people interested/skilled • Who pay the initial costs • Who pay the long term operation

Equipment • Before buying equipment, make sure main purpose of network is clear and that the funding for installation, operation and training is in place, ideally start with training • For a new network, the main purpose will be location and magnitude and there might not be a need for sophisticated instrumentation, simple and reliable is the main aim • There will often be a push to install a broad band network, however accelerometers, SP seismometers or even geophones might do as well in order to get unclipped data, at a substantial lower price • Make sure enough money is spent on processing equipment

Software • It is important to use a free, well tested software that can do all the routine operations required by the network of which the most important is: Local and global locations, calculate common magnitudes, correct for instrument response and store the data in a systematic way. • The software should be able to be used on common computer platforms and use common formats. • Training should be available

SEISAN • SEISAN has been used for more than 20 years • Has been an important tool for sharing software and data • Used by more than 100 institutions, mostly in Europe, Africa, Latin America and India • 27 % of agencies reporting to the ISC use SEISAN and report in SEISAN format • Used by some instrument companies when a low cost solution is required by the customer • Used at international training courses like UNESCO-GFZ

Open data Too many individuals/and or organizations refuse to share data. The purpose is not to use the data but to get the power of controlling the data • • Not sharing will prevent cooperation • When planning and/or upgrading a network, a firm requirement should therefore be that the data is open

Workshops, not always ideal There are every year several seismology workshops and training courses. Some will serve the purpose of useful training, however for many, the participation seems mainly to motivated by the event itself. Some participants do not have the relevant background and will never work with the material again. Our experience is that no more 20 % of participants get direct use of the training. The more efficient training is using the participants own equipment and data.

Summary NMSOP recommendations for setting up a new network • To ensure long term maintenance and get lowest possible operational cost, the user should build as much as possible of the network himself • Study and visit other networks to learn the requirements • Decide on communication • Decide on type of network based on communication possibilities • Decide on integration of all components into one system • Make sure processing and data storage systems are in place (including training) before network installation is finished From New Manual Of Seismological Observatory Practice (NMSOP)

Uganda • Uganda got a UNESCO grant to build one analog three component analog station in Entebbe (ca 1980). • At this relatively high noise site, only few events could potentially be detected from the seismic area in the rift valley. • For the same money 4 MEQ 800 was bought instead, and 3 were installed, one was used as a spare. Local operators took care of the field stations. It was now possible to locate events and calculate magnitude. • This network operated successfully more than 15 years (later with addition of digital stations) by the Geological Survey of Uganda.

Tibet Seismicity near Lhasa 2008. The number of recordings for this period is more than 2128, while 1837 of them fall in the above region. The stations are marked with blue. A typical installation of each of the four current stations of the network. The sensor is to the left (small gray box), and the digitizer to the right.

Money is not all but it helps Small Gulf country get a turnkey system with broad band stations, high technology. Local company does all maintenance so network operates well. Instrument company provides simple software for local earthquakes, but many events are global so they are located inside country, suddenly the country has a lot of seismicity ! Clear example of too little money for training, compared to money for instruments.

How it can go wrong 1 • Large company delivers network • Licensed software installed, partly automatic processing • License runs out, no money for renewal • Not enough local training to take over operation with free software Lesson: Too much money for too sophisticated equipment and no contract to make sure the data actually comes out the other end finally processed. The customer should have started with the processing-people end and then seen what was needed in terms of equipment. And the instrument company might have painted their equipment and automatic processing it bit too rosy. Automatic processing should not be used with a new network operated by people with little training.

How it can go wrong 2 • International organization installs station in country with little seismological expertise • Data is transmitted directly to donor country and local access to the data is either, nonexistent, offline or unreasonably complicated • If software is provided for local use, it is often quite useless and made to ‘have done the duty’. The result is that local use of the data is very limited. Lessons learned: If outside organizations offer to set up a local stations with real time transmission to their organization, make sure the operation includes: • Complete local access to data • A simple way to find and extract out local data • Clear guidelines and training in how to use the data and be able in a simple way of integrating it with possible local data.

Using existing local and regional resources • Before installing more equipment, make sure all existing data is fully used. • Some countries too small or have too few stations to make reliable locations • Regional cooperation will then help to: – Provide more data – Provide help in operation and processing

East African Stations, 1992 Stations and countries cooperating in seismology in 1993 Dindi et al, 1995, SRL 85, 364

Central America Alvarenga et al, 1998, SRL 69 Seismic stations (138) in Central America, May 1998

Conclusion • Find local people interested in the network, provide initial training • Make sure current resources are used before making new investments • Only buy the equipment needed for the basic goal, avoid equipment and software that require yearly license fees or expensive upgrades, avoid automatic processing systems • Use local people for installation • Cooperate regionally if possible • Make sure funding for operation and training is available