Statistical and geostatistical analysis related to geographical parameters for spatial and temporal representation of rainfall in semi-arid environments: the case of Algeria

Abstract

The economic challenges related to the fields of agriculture and industry led us to adopt the best suited method to represent the rain on the spatial and temporal plan especially in areas characterized by heterogeneous rainfall distribution additionally to drought periods. The methods of analysis and estimation of rainfall, using a number of tools (statistics, geostatistics and digital mapping), provide the opportunity to represent the average inter-yearly rainfall fields in the eastern high plateaus region of Algeria. In this study, an approach was proposed for yearly rainfall characterization. Data series for the period 1986–2007 were collected from 65 rain-gauging stations. This approach is based on two combined methods (geostatistic and multiple linear regression) including direct relationship between rainfall and geographical parameters (longitude, latitude and altitude). Statistical analysis indicates that the annual rainfall values ranges from 127 to 752.2 mm and that their distribution is Platykurtic. Results show that yearly rainfall structure obeys mainly a north/south gradient, and latitude is the most influential geographical parameter with a coefficient of 261.25 contrary to the longitude (17.06) and altitude (0.04) which have a non-significant effect on precipitation. In addition, other factors such as vegetation, temperature and air mass movement affect negatively the rainfall structure. Moreover, the map of rainfall indicates that the rain bands ranging from 300 to 400 mm dominate 58 % of the total study area whereas rain bands greater than 400 mm occupy 37 % of the total study area.