Drought Proneness Analysis of Southern Saskatchewan Province Using Markov Chain Model

The southern Saskatchewan region has been recognized for persistent droughts due to its continual rainfall deficiency. To estimate drought severity, a first-order and five-state Markov chain model is used in this study. Long-term daily precipitation records of three stations (Broadview, Last Mountain CS, and Swift Current CDA) were considered for the analysis. The Standard Precipitation Index (SPI) was used to classify the drought severity into five different classes (no drought, mild drought, moderate drought, severe drought, and extreme drought) to develop Markov models. Transition Probability Matrices (TPM) were calculated for each of the stations, and the chances of drought occurrence were estimated from the steady state matrix. Expected drought duration, mean first passage time of drought, and mean recurrence time of drought were also calculated from TPM to understand the drought characteristics in the long run. The results indicate that mild drought has more than 35% chance of occurrence, and the expected drought duration of extreme drought can exceed over a month. Furthermore, for extreme drought, the mean recurrence time and the mean first passage time can be varied from 102.09 to 168.20 months and 14.47 to 20.20 months, respectively. The Markovian approach is an efficient method to estimate and understand the drought characteristics, providing crucial information in agricultural sectors, planning, and other related professional fields to adapt and plan accordingly.