Skip to main content

Modeling Dryness Severity Using Artificial Neural Network at the Okavango Delta, Botswana

  • Authors (legacy)
    Corresponding: Jimmy Byakatonda
    Co-authors: Byakatonda J., Parida B.P., Kenabatho P.K. and Moalafhi D.B.
    Download PDF
  • gnest_01731_published.pdf
  • Paper ID
    gnest_01731
  • Paper status
    Published
  • Date paper accepted
  • Date paper online
Abstract

Water balance studies in the Okavango Delta indicate that more than 90% of inflow into the Delta is lost through evaporation. This coupled with high climatic variability threatens the ecohydrology of the Delta. Trends indicate decreasing rainfall amounts and increasing temperature at the area of the Delta. The main aim of this study was therefore to investigate long term trends and variability in rain onset, cessation, number of rainy days and their impact on the dryness index at the Delta. The impact of the above variables is expressed through the standardized precipitation and evaporation index (SPEI) quantified by aggregating the climate water balance and fitting monthly series to a generalized logistic distribution using L-Moments. The SPEI, determined at windows of different time scales of one, three and twelve months, provided an extensive evaluation of dryness severity and its impact on this sensitive ecosystem. Rain onset and cessation dates were generated from cumulative pentad rainfall–evapotranspiration relationships. Analysis of climatic data showed mean rain onset occurring in November and ceding in March with average of 44 rainy days between 1970/71 and 2013/14. The results revealed a decrease in the number of rainy days at a rate of 0.16 days/yr and of the duration of the rainy season at 0.25 days/yr with high variability. Annual rainfall was found to decrease at the rate of 1.60 mm/yr with 6.8% probability of failure in rainfall onset. Analysis further revealed that both extreme dryness and wetness are rare phenomena with probabilities of less than 1% and near normal conditions for 67% of the time for all SPEI time scales. Although gradual increase in dryness in the Delta is attributed to high climatic variability, simulations undertaken using Artificial Neural Networks did not predict any major changes in the next five years. However, vulnerability to severe droughts is not completely ruled out because of the high variability in rainfall and of the location of the Delta in a semi-arid zone. 

 

Copy to clipboard
Cite this article
Parida, B. et al. (2016) “Modeling Dryness Severity Using Artificial Neural Network at the Okavango Delta, Botswana”, Global NEST Journal, 18(3). Available at: https://doi.org/10.30955/gnj.001731.