The ANSTO CMDL Radon Program at MLO
S. Whittlestone
Australian Nuclear Science and Technology Organization, Menai, NSW 2234, Australia
D. Kuniyuki and S. Ryan
NOAA Mauna Loa Observatory, Hawaii
Radon is measured at the Mauna Loa Observatory, Hawaii (MLO) to characterize air masses. It is produced from the radioactive decay of radium in soils and decays with a half-life of 3.8 days, so that an air mass that has been away from land for longer than 2 weeks will contain very little radon. There are two uses for radon data: (1) as one of the parameters for selection of baseline air samples and (2) as a tracer to test the accuracy of global scale air transport models.
The latest version of the Australian Nuclear Science and Technology Organization (ANSTO) radon detector, installed at MLO in 1994, has proved to be mechanically and electronically reliable, and the calibrations were steady over a long period. The seven calibrations in 1997 had a standard deviation of 5%. The only instrumental failure was caused by a lightening strike in October 1997. Fortunately this destroyed only the automatic calibrator that was scheduled for replacement in any case.
Data are recorded directly into a data logger with a 3-month memory capacity. This logger is queried every 6 hours and the data is transferred to a computer that runs Netbeui and TCP/IP protocols under DOS for Workgroups. A program written in the BASIC language writes data to the hard drive on the local computer and to a network-shared directory residing on the server at the observatory. This allows data to be accessed in near-real time through the network by FTPing to the server or doing a map network drive from a Windows 95 or NT system. In either case a password must be given to gain access to the data directory.
The major instrumental problem has been in data retrieval. Some event, probably electrical interference from equipment elsewhere at MLO, periodically crashes the computer and sends faulty communication codes to the logger that result in corrupted data files. More robust software installed in November 1997 has reduced the effect of these events, but further work is needed.
Figures 1 and 2 show hourly radon concentrations at MLO in 1996 and 1997. At the compressed time scale of the graphs, the diurnal variation results in a band of varying width superimposed on the longer term variation of tropospheric radon from Asia. In both years, the spring and fall experienced periods of high radon with proportionally little diurnal change, whereas in summer low levels of radon with proportionally greater diurnal variation prevailed.
Fig. 1. Radon concentrations at MLO in 1996.
Fig. 2. Radon concentrations at MLO in 1997.