On the method of estimating the event of unusually high values in the radiation monitoring data
Creators
- 1. Japan Atomic Energy Research Inst., Tokai, Ibaraki. Tokai Research Establishment
Description
The method was discussed for estimating a comparatively rare occurrence of monitoring data much greater than the average in the field of radiation protection. The features of and the mutual relationships among four types of probability distributions, i.e., normal, log-normal, double-exponential and log-double-exponential, were presented, and their application to various monitoring data of radiation protection was shown. For graphical analysis of empirical distributions the followings were concluded: 1. The upper probabilities are considerably affected by the method of calculating the quantiles and the method of logarithmic transformation of a stochastic variable after a fixed value being added to or subtracted from it to obtain the better straight-line fit on the probability paper. 2. A double-exponential distribution provided a good fit to the annual dose maxima, while a log-double-exponential distribution did to the radioactive contamination maxima in a fixed period of time or in a fixed region of space. 3. The monitoring data arranged in the order of their magnitude were expressed by an exponential function of the order for dose data and by a power function of the order for radioactive contamination data. (author)
Additional details
Publishing Information
- Journal Title
- Hoken Butsuri
- Journal Volume
- 15
- Journal Issue
- 2
- Series
- Hoken Butsuri.
- Journal Page Range
- 101-109
- ISSN
- 0367-6110
INIS
- Country of Publication
- Japan
- Country of Input or Organization
- Japan
- INIS RN
- 13660121
- Subject category
- S99: GENERAL AND MISCELLANEOUS; S61: RADIATION PROTECTION AND DOSIMETRY;
- Descriptors DEI
- CONTAMINATION; DISTRIBUTION FUNCTIONS; PROBABILITY; RADIATION DOSES; RADIATION MONITORING; RADIATION PROTECTION; STATISTICAL MODELS
- Descriptors DEC
- MATHEMATICAL MODELS; MONITORING