ATMOSPHERIC TRITIUM ANALYSIS. Technical Progress Report No. 3

Technical Report · Mon Nov 27 00:00:00 EST 1961

OSTI ID:4813034

Barrett, E W; Huebner, L

The sequence of measurements of tritium in Chicago precipitation was extended into the winter of l960 and l96l. Precipitation sampling was augmented near the end of the period by samples of water vapor condensed directly from the ambient surface air by means of mechanical refrigeration equipment. The tritium contents of such condensed vapor samples agreed to within the standard error of the measuring technique with the tritium contents of simultaneously-occurring precipitation. This observation, if verified by further data, indicated that the rate of tritium equilibration between falling precipitation and ambient water vapor is fairly rapid. The continuous sequence of tritium data, when smoothed by taking 100-day running means, showed an asymptotic decay superimposed upon an annually-periodic cycle with peaks in the early spring. The first study assumed a two-reservoir model of the atmosphere; the stratospheric reservoir discharges into the troposphere with a rate coefficient which is a periodic function of time with a spring maximum, while the troposphere discharges into the hydrospheric sink with a constant rate coefficient. The rate coefficient for the troposphere was found to be l7.4, corresponding to a mean residence time for water vapor of 2l days. The periodically varying stratospheric storage time was found to have a spring-minimum effective value of 0.42 year, or 153 days, and an autumn maximum of 3.28 years. The annual-mean value for this residence time is l.l7 years. The model gave an excellent fit with the smoothed data for the first year;, systematic deviations appeared during the second. These deviations were explainable by the assumption of two stratospheric reservoirs, a low-latitude reservoir with a storage time long compared to one year and a high-latitude reservoir with the storage time as given abeve. It was found, on the basis of the calculated parameters of the tworeservoir model, and on the assumption of a mean value of 3 cm for the tropospheric precipitable water, that the total tritium content of the stratosphere at the beginning of November 1958 was 27.6 kg while that of the troposphere was 255 gm. The stratospheric burden was reduced to l.66 kg by the end of October 1960 by transport to the troposphere and radioactive decay. The annual-mean rate of exchange of water vapor across the tropopause by all processes was calculated to be 7.6 x l0/sup -3/ cm of precipitable water. It seemed likely that only a small fraction of the water vapor in the stratosphere enters as vapor in tropical latitudes; most of the water vapor is transported toward high latitudes by eddy-exchange in the troposphere and enters the stratosphere by eddy-mixing processes north of the subtropical high-pressure belt. An attempt to deduce the parameters of a three- reservoir model through application of the method of least squares is described. Two trials of the problem on an IBM 704 computer failed to converge after 69 and 20 iterations, respectively. It is believed that the slowness of convergence was the result of a set of poor first estimate of the parameters. (auth)

• Research Organization: Chicago. Univ.
• DOE Contract Number: AT(11-1)-636
• NSA Number: NSA-16-006369
• OSTI ID: 4813034
• Report Number(s): TID-14433
• Resource Relation: Other Information: Orig. Receipt Date: 31-DEC-62
• Country of Publication: United States
• Language: English

Related Subjects

CHEMISTRY
AIR
ATMOSPHERE
COMPUTERS
DECAY
DIAGRAMS
DIGITAL SYSTEMS
ERRORS
GEOPHYSICS
IBM 704
LABORATORY EQUIPMENT
LIQUEFYING
LOW TEMPERATURE
MATHEMATICS
MEASURED VALUES
METEOROLOGY
MIXING
NUMERICALS
PERFORMANCE
PRESSURE
RADIOACTIVITY
RAIN
RATE METERS
REDUCTION
SAMPLING
SEASONS
STABILITY
STANDARDS
STORAGE
STRATOSPHERE
TRANSPORT
TRITIUM
TROPOSPHERE
VAPORS
VESSELS
WATER