A study on the global detection capability of IMS by kellena96


									               A study on the global detection capability of IMS for all CTBT relevant xenon
                                                                                                                                                      Réal D'Amours1 and Anders Ringbom2
                                                                                                                                             1- Canadian Meteorological Centre (CMC), Environment Canada, Dorval, QC, Canada
                                                                                                                                                 2- Swedish Defence Research Agency (FOI), SE-164 90, Stockholm, Sweden

1.Introduction                                                                                                                                                     4. Dispersion model

                                                                            Detection limit (mBq/m3)

  Previous studies of the measurement capability of

  the radioxenon part of the International Monitoring
  System (IMS) for CTBT monitoring have all been                                                                    0.1
                                                                                                                                                                   It was impractical to calculate SRS fields to form a full
  made using simple assumptions with regard to the                                                                 0.01                                            climatological database. In order to sample a representative
  detection limits of the measurement systems and                                                      -90   -45          0        45   90
                                                                                                                                                                   spectrum of meteorological conditions, it was decided to select
  without taking the existing background into account.                                                        Latitude (degrees)
                                                                                                                                                                   a sequence of 7 days, 3 days apart. The inverse model was
  Furthermore, all earlier studies have been done for           Fig. 1: Measured 95:th percentiles of the atmospheric activity of                                  executed for days 3-6-9-12-15-18-21 in July and December.
      Xe only. The recent development of sensitive              133Xe as a function of latitude for the IMS sites used in this study.
                                                                                                                                                                   Calculations were repeated for 4 years 2005-2006-2007 and
  radioxenon systems and the build-up of the network            The results illustrate the well known fact that the xenon
                                                                background is higher at the northern hemisphere. The highest                                       2008. The simulations ran 14 days backward.
  within the International Noble Gas Experiment                 detection limits were obtained for stations close to the isotopic
  (INGE), has substantially increased the possibility to        produciton facilities in Chalc River, Canada, and Fleurus, Belgium.
                                                                                                                                                                   This resulted in 56 14-day simulations to estimate the coverage
  perform a more realistic study of the network                                                                                                                    for July and December. The SRS fields were calculated for the
  performance in the case of a nuclear explosion.                For the metastable states, and for 135Xe, typical critical
                                                                                                                                                                       Xe, 131mXe, 133mXe, and 135Xe isotopes, for every 3-hour
                                                                 limits were used for the majority of stations. In a few                                           interval, 14 days backwards in time.
The goal of the study presented here was to:                     cases the same method as described for 133Xe was
                                                                 used. The detection limits for 131mXe and 133mXe were,                                            The fields were combined with the assumed source term to
• Use the data set presented in [1] to define detection          with a few exceptions, in the range 0.06 -0.1 mBq/m3,                                             calculate a concentration at the station. The first statistic
  thresholds for 133Xe, 131mXe, 133mXe, and 135Xe for            while a typical detection limit for 135Xe was 0.3 mBq/m3.                                         calculated was the frequency at which a given location would                                                                                            131mXe, 133mXe, 133Xe
                                                                                                                                                                                                                                       Fig. 3: Average number of stations (including variance) detecting a release of                              and 135Xe assuming
  all IMS sites. The thresholds should be calculated             For the remaining IMS sites, the detection limit was                                              have been detected by that station using it's estimated             the source term described in Section 3.
  taking the local radioxenon background into                    estimated based on the general site location.                                                     threshold during the span of the 14 days. Then the 39
  account.                                                                                                                                                         frequency fields were combined into a single field showing the

• Use the CMC global Lagrangian dispersion model in
                                                            3. The radioxenon                                                                                      number of stations that would have seen the location at least
                                                                                                                                                                   once during the span of the 14 preceding days. Finally the
  inverse mode to calculate Source Receptor
  Sensitivity (SRS) fields for each station in the IMS       source term                                                                                           average and the variance of the number detecting stations
                                                                                                                                                                   was calculated for the 28 simulations in July and the 28 in
  network.                                                                                                                                                         December. The last calculations were also repeated for 133Xe
                                                            The activities of the different radioxenon isotopes                                                    and 135Xe using a minimum frequency of 4 detections for each
• Combine the SRS fields, the station dependent             following a nuclear explosion are strong functions of                                                  station.
  thresholds, and a plausible radioxenon release term       time (see Fig. 2), and the isotopic ratios for a prompt
  to estimate the global isotope dependent radioxenon
  detection capability of the network. The study should
                                                            release will be very different compared to a release
                                                            after one day. Here we assume a source term based                                                       5. Results
  include seasonal variations, and be conducted for            on
  both the 39 station noble gas (NG) network currently      a 10% release of the activity produced by a 1 kt                                                        The results of the calculations are displayed in Fig. 3 and Fig.   Fig. 4: Average number of stations (including variance) detecting a release of      133Xe   and 135Xe four times, assuming the
  being built and the 79 IMS station network (only          plutonium charge after 2.78 days, when 133Xe reaches                                                    4. The coverages are quite similar for133Xe, 131mXe, and           source term described in Section 3.
  results for the NG network are reported here).            maximum. This results in the following activities:                                                          Xe, while the network detection capability is, as
                                                                                                                                                                    expected, considerably less for 135X because of its very short
                                                                                                                                                                    half life. The detection capability is higher in winter for both
2. Calculation of site-                                         131mXe
                                                                                                                                                                    hemispheres, because of increased meteorological activity in
                                                                                                                                                                    that season, and faster air streams. The detectability is
 dependent detection                                                                                                                                                                                                                              • An estimation of detection capability of the NG network has been done using
                                                                133Xe                                        1.06e+15                   Bq/kt                       higher in the northern hemisphere likely because of the
                                                                135Xe                                        3.91e+14                   Bq/kt                       higher number of stations, and very low in the equatorial                     realistic estimates of the detection levels at stations based on radioxenon

 limits                                                                                                                                                             regions, especially in the mid Pacific, because of fewer
                                                                                                                                                                    stations and reduced atmospheric transport in the tropics.
                                                                                                                                                                                                                                                  backgrounds statistics.

                                                                                                                                                                    The detectability is also low on the Tibetan plateau in July.                 • The CMC Lagrangian dispersion model was used to calculate a detectability
The observed 133Xe background at 17 IMS sites was                                                                                                                                                                                                 statistic: the number of stations in the NG network detecting the location at least
fitted to normal or lognormal distributions as described                                                                                                            Interestingly the variance of the number of detecting stations                once during the preceding 14 days.
in [1]. The lognormal distributions were folded with the                                                                                                            seems to be higher over the regions were the gradient of the
concentration uncertainty as a function of                                                                                                                          average is large, and tends to be lower over the regions                      • The statistic indicates that the network capability is higher in the mid latitudes
concentration. In cases when the distribution obviously                                                                                                             where the average reaches its maximum. This is more                           and the polar regions. There is a strong seasonal effect, detectabilty being higher
not was lognormal, the result from data itself was                                                                                                                  apparent in December in the Northern Hemisphere, over the                     in winter.
    used,                                                                                                                                                           Atlantic. Again this is probably attributable to the nearly
without fitting procedure. The resulting distribution was                                                                                                           constant high meteorological activity over that area in                       • The coverage is very low in the equatorial regions, especially in the Pacific.
used to define a detection limit at the 95th percentile                                                                                                             December that ensures a rapid transport to the European NG
for each site (see Fig. 1).                                 Fig. 2: Radioxenon activities as a function of time following fission of 239Pu                          stations.                                                                                                                135
                                                            (PBq/kt). The calculations take into account the full decay chains [2].The
                                                                                                                                                                                                                                                  • The coverage is very low for                 Xe (0-1 stations), but fairly similar for the other
                                                            vertical green line marks the time chosen for the source term in this work.                                                                                                           isotopes (0-15 stations).
                                                                                                                                                                    As expected, the detectability is lower when using a threshold
                                                                                                                                                                    frequency of 4 detections per stations (only 133Xe and 135Xe        [1] A. Ringbom et. al., “Characterization of the global distribution of atmospheric radioxenons”, ISS09, Vienna, 2009.
                                                                                                                                                                    are shown) . However the patterns remain similar.                   [2] L.-E. De Geer, Rep. FOI-R—2350—SE.

To top