Open AccessUsing Artificial Neural Networks and SPI Measure Techniques to Forecast the Risk of Drought in Iraq and Its Impact on Environment
Author(s) -
Renas A.A. Nader,
Aras J.M. Karim,
Mohammed L. Hussien
Publication year - 2018
Publication title -
govarî zankoy geşepedanî miroyî
Language(s) - English
Resource type - Journals
eISSN - 2411-7765
pISSN - 2411-7757
DOI - 10.21928/juhd.v4n2y2018.pp69-77
Subject(s) - desertification , arid , agriculture , economic shortage , precipitation , water scarcity , environmental science , artificial neural network , geography , population , climatology , physical geography , meteorology , ecology , demography , biology , computer science , geology , linguistics , philosophy , archaeology , machine learning , government (linguistics) , sociology
The world suffers from drought, which has a negative impact on human, economic, social, cultural and tourism fields. As science progressed and developed, several ways of reducing drought were found. This phenomenon is also called (aridity and infertility, and water retention), it means a severe shortage of water resources due to low precipitation and low rainfall over a specific normal period time, which are causing heavy losses in agricultural production, and the occurrence of disasters and human calamities such as starvation, and it is forcing some population to emigrate collectively. The artificial neural networks (ANN) and the Standard Rain Index (SPI) were used in the analysis of the rainfall for all Iraqi governorates for the period 1991-2016 monthly. This study shows that the best model of the neural network is [19-3-1] according to AIC to forecast the amount of rainfall, and that the Iraqi provinces over next 10 years are exposed to a different behavior of climate between moderate dry and average humidity, and increase the area of desertification.