A Markov chain model for daily precipitation occurrence in South Korea

✍ Scribed by Sung-Euii Moon; Sang-Boom Ryoo; Jae-Gi Kwon

• Publisher: John Wiley and Sons
• Year: 1994
• Tongue: English
• Weight: 480 KB
• Volume: 14
• Category: Article
• ISSN: 0899-8418
• DOI: 10.1002/joc.3370140906

✦ Synopsis

Abstract

The Markov chain properties of daily precipitation occurrence, namely, the dependency of the daily precipitation on that of the previous day, are studied based on the daily precipitation data for 30 years (1956–1985) at 14 stations in South Korea. The daily precipitation data at each station were classified into the wet and dry state, and transition probabilities between daily precipitation of two successive days were computed. Then the Markov chain properties and various aspects of daily precipitation occurrence from the Markov chain properties were investigated. The results are as follows: the transition probability of two successive wet days for 30 years at the 14 stations is 0.51, and the statistical tests show that the transitions of daily precipitation occurrence in South Korea can have the Markov chain property and be stationary in time, except at Ullung‐do, Seoul, Kangnung, and Mokpo, but heterogeneous in space. The n‐step Markov chain analysis shows that the ‘memory’ of the daily precipitation occurrences at 14 stations remains about 2–3 days. The weather cycle is 6.59 days.