Untitled Document

Enhanced infrasound monitoring stations for CTBT verification

Douglas R. Christie

Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia

The signing of the Comprehensive Nuclear-Test-Ban Treaty (CTBT) on 24 September 1996 has led to a rapid development in infrasound monitoring technology for Treaty verification. A 60-station global infrasound monitoring network is under construction and nearing completion. This network is designed to reliably detect infrasonic signals from a 1-kiloton atmospheric nuclear explosion at two network stations. Global three-station detection capability is desirable since this would greatly enhance network reliability, lower detection thresholds and improve location estimates. Work carried out at the ANU during the last year on monitoring array design and background noise suppression suggests that the performance and reliability of the infrasound component of the International Monitoring System (IMS) can be enhanced considerably.

A detailed investigation of the spatial correlation of infrasonic signals from distant explosions has shown that the low degree of signal correlation between elements in many existing monitoring arrays may limit detection capability for regional and distant explosions. A new technique based on the Mack and Flinn coherence model has been developed to calculate the complete azimuthal variation of the array-averaged correlation coefficient for arbitrary array configurations. This azimuthal pattern provides a unique array characteristic that can be used directly as a measure of array performance. This technique has been applied to a wide variety of arrays, ranging from 3 to 9-element arrays, in an attempt to find an optimal design for an IMS infrasound monitoring array. The results of this investigation show that the detection capability of IMS stations with a small number of array elements for distant explosions may be marginal when the array aperture exceeds 1 kilometer. These studies also show that the performance of many existing 7- and 8-element arrays in the IMS is not optimal.

The performance of these arrays decreases rapidly at high frequencies and the sensitivity may exhibit significant azimuthal variation in the monitoring passband. A careful examination of a large number of possible array configurations shows that an optimal array for nuclear explosion monitoring can be achieved by using a 9-element array configuration arranged in the form of a small aperture (350 m) centered triangle sub-array located at the center of a larger aperture (1 km) pentagon array. This array design is recommended for future use at IMS infrasound monitoring stations.

Wind-generated noise in the primary monitoring passband (0.4 to 1.2 Hz) is a serious problem at many infrasound stations. Stations located in open exposed areas are often subject to unacceptably high levels of background noise. Currently used wind-noise-reducing pipe arrays provide a significant reduction in background noise levels, but the degree of wind-noise-reduction may not meet CTBT verification requirements, especially during the daytime when the boundary layer winds are coupled to the surface. In order to meet essential monitoring requirements, wind-noise-reducing systems need to provide at least two orders of magnitude reduction over that provided by currently used pipe array systems.

We have therefore developed a new wind-noise-reducing system that is capable of effectively eliminating wind noise in the monitoring passband at most infrasound monitoring stations. This system is based on the use of a series of screens which effectively degrade turbulent eddies and lift the turbulent boundary layer over the sensor inlets. This device is referred to as a turbulence-reducing enclosure. A large number of designs have been tested. Initially, these enclosures were constructed as open enclosures with concentric porous walls with overlapping deep serrations along the top of each wall inclined away from the center of the enclosure. These structures provided more than two orders of magnitude noise reduction. Thus, these structures may be used in some cases with existing pipe arrays to achieve acceptable background noise levels. The latest version of the noise-reducing enclosure is, however, much more efficient. This version is constructed in the form of a closed enclosure with a porous screened roof, internal baffles, and multiple interior chambers.

This highly efficient noise-reducing system has been tested with both a small 6-port pipe array and with a single-inlet port system located at the center of the enclosure. This system provides more than 4 orders of magnitude noise reduction (see Figure 1) in winds of up to at least 6 m/s at a height of 2 m above the surface. It is interesting to note that the single inlet port system is more efficient than the 6-port pipe array at frequencies above 0.7 Hz. This new system can therefore be used in some cases as a stand-alone system that does not require a pipe array. We recommend, however, that this new noise-reducing system should be used at IMS stations in conjunction with existing pipe arrays in order to completely eliminate wind-generated background noise in the monitoring passband.


Figure 1.  Power spectral density of background noise data recorded simultaneously inside and outside a turbulence-reducing-enclosure during typical daytime wind conditions at IMS monitoring station IS07 Warramunga.