Temporal Distribution and Drought Characteristics of Streamflow Evaluated using Standard Streamflow Index and Markov Chain Approach-converted
Authors: Kee An Hong¹ and Jer Lang Hong
Hong & Associates, Kuala Lumpur, Malaysia
SNS Network, Kuala Lumpur, Malaysia
Corresponding author: Abdul Wahab Mohammad
E-mail: firstname.lastname@example.org , email@example.com;
A three state Markov chain was used to examine the patterns and distributions of monthly runoff data based on the standard streamlow index (SSI) of two basins in Peninsular Malaysia, i.e. Bernam and Kesang which are located in a dry and a comparatively wet area. The conditions of drought severity were assessed using the SSI. The droght conditions are classified into 3 states, dry, normal , wet according to the SSI at each month for time scales of 6 and 12 months. A month is said to be dry if the SSI< -1, normal if the SSI >-0.99 and <0.99 and wet when SSI >1.Based on the three conditions and statistical techniques applied , it was observed that the expected lengths of dry , normal and wet months are 5.8,10.8 and 4.1 for Bernam and 5.2,10.4 and 4.8 for Kesang for the 6 month SSI.For 12 month SSI, the values for Bernam are 9.3,17.1 and 5.6 and 6.3,15.5 6.2 for Kesang. The weather cycle for Bernam and Kesang are20.7 and 20.5 for 6 month SSI and 32 and 28 for 12 month SSI. It is noted that the two basins although are wet and dry with different mean annual runoff ,results show that the runoff distributions and patterns of the two basins are quite similar as theyare located in the same general hydrological region experiencing similar wet and dry seasons.
Keywords: SSI, SPI, Markov chain, transition probability, dry spell, weather cycle, WS