Stochastic characteristics of solar irradiation and temperatures in Botswana: ARIMA models

J. Prakash, E. M. Lungu, P. K. Jain

Research output: Contribution to journalArticle


This paper presents bivariate models that relate solar irradiation to extreme temperatures for Sebele, Botswana. Autocorrelation analysis revealed that the solar irradiation series is stationary for d = 2 and D = 0, and the extreme temperatures series are stationary for either d = 1 and D = 1, or d = 0 and D = 1, 2... N. A lag of three months is found between the peaks of the differenced series of solar irradiation and maximum temperatures, whereas there is a lag of at most one month between the differenced series of solar irradiation and minimum temperatures. Analysis of the noise component revealed that the bivariate processes under consideration behaved either as ARIMA processes of order (0,1,1) × (0,1,1)12 or as MA processes of order (0,1,1)12. However, where non-seasonal influences are present, i.e., in the case of model (0, 1, 1) × (0, 1, 1)12, these effects exhibit month to month memory only. The relationships found for Sebele can be applied to estimate extreme temperatures at other locations in the region, where similar climatic conditions prevail.

Original languageEnglish
Pages (from-to)91-101
Number of pages11
JournalInternational Journal of Meteorology
Issue number359
Publication statusPublished - Apr 1 2011


All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this