The knowledge of natural risk of occurence of drought period represents a necessity for the foundation of afforests technologies and silvicultural operations. The idea of synthesis is in the book of H. Walter (1974) and is based on the using of climatic diagrams type Walter-Leith for the delimitation of drought climate of Romania (Doniţă, 1967; Barbu şi Cenuşă, 1975). Walter (1974) considers that the period in wich the precipitation curve at P = 2T scale is under the temperature curve is a drought period and the period and the precipitation curve at P = 3T scale is under the temperature curve is an aridity period. The delimitation of drought zone on the base of climatic diagrams type Walter-Leith (P < 2T) and aridity zone (P < 3T) in function of lenght of drought period or aridity period was made on basis of the analysis of the means from 27 weather stations. The greatest correlation was found between the number of drought (aridity) days and the annual mean precipitation of station (r = -0.82). The maximum number of days with drought is regitered in the regions with precipitations smaller than 400 mm/year and is greater that 50 days of drought and 150 days of aridity and the minimum in the regions with Pma grater than 600 mm/years. Practically, in these regions with Pma greater the drought is missing and the number of aridity days is lower than 50 days/year. The multiple regressions between the number of drought days (aridity) and the geographic (altitude, latitude, longitude) and climatic parameters (Tma, Pma were established:
Nu = 245 - 0.015Alt - 3.52Lat + 0.67Long - 0.37Pma + 10.6Tma (R multiple = 0.88, R square = 0.78)
Ns = 102.4 - 0.15Alt - 0.034Lat + 1.92Long - 0.36Pm + 13.89Tma (R multiple = 0.915, R square = 0.838)
The multiple correlation coefficients calculated have very significant values. This permit the graphical representation of number of drought days (aridity) in relation with 2-5 geographical parameters using geostatistical analysis.
Keywords: drought, aridity, modelling