As noted in the discussion of wave propagation in shallow water and of wave runup, the local conditions at a tide gauge location or at any other coastal site will greatly influence wave amplitudes at those sites; furthermore, tide gauges tend to underestimate the amplitudes of shorter - period waves in particular. Experience (as in the 1960 Chile tsunami, when lives were lost in Hawaii and Japan because insufficient evidence from tide gauges was available to convince emergency managers that a major tsunami event was in progress) has shown that tide gauges on some of the very small coral atolls in the southern Pacific, on the paths of tsunamis propagating from Chile and Peru to Hawaii and Japan (or vice versa) only record very small wave amplitudes even though the tsunamis are very energetic: this may be because of sheltering of the tide gauges by the coral reefs and the fact that the islands are very small compared to tsunami wavelengths in the surrounding deep water, and therefore do not disturb the passage of the waves. More generally, tide gauges provide little reliable information on the magnitudes of tsunamis and hence tsunami warnings consist merely of the information that a tsunami event is in progress, not its magnitude: this is another contributor to the false alarm problem. In future, more reliance will be placed upon deep-water buoy systems.