Comparison of two methods of erosive rains determination

摘要

Number of erosive rains, kinetic energy of erosive rains and factor of erosive efficiency of rains according to the USLE methodology were assessed by two methods of erosive rains determination. The first method (VAR1) defined erosive rains by intensity ≥ 0.4 mm· min-1; total ≥ 12.5 mm and the second method (VAR2) by intensity ≥ 6 mm· 15 min-1; total ≥ 12.5 mm. Database contained one minute precipitation data from four automatic stations in the Czech Republic for the period of 2000-2005. Two-way analysis of variance (ANOVA) showed a statistically highly significant difference between the annual number of erosive rains determined by the two methods. The rains simultaneously complying with two following criteria (30 min intensity lower than 15 mm·h?1 and sum of 40 mm) were not generally classified as erosive rains according to VAR2. The number of erosive rains determined by VAR2 most often reached 40 to 50% of VAR1 results. Two-way ANOVA proved highly significant differences between the kinetic energy values for the erosive rains determined by VAR1 a VAR2. According to VAR2 the rains with kinetic energy lower than 3 MJ·ha ?1 are generally not considered as erosive rains. The results of kinetic energy of the erosive rains determined by VAR2 most often reached 60 to 70% of VAR1 results. Two-way ANOVA has not proved a statistical difference between annual values of R factor of erosive rains determined by the two methods. According to VAR2 the rains with R factor lower than 5 are in general not included into annual R factor value. The results of annual R factor values of erosive rains determined by VAR2 are about 25% lower than the results of VAR1. Correlation between number of erosive rains, kinetic energy of erosive rains and annual R factor value assessed by both methods showed a statistically significant relationship. The conversion formulas between results of the two methods (VAR1 and VAR2) were derived by linear regression. As conclusion we can state that when using present automatic stations in R factor analyses, we have to be aware of overestimating the erosivities compared to historical data based on ombrograms, where only low temporal resolution data were available.