Effects of four storm patterns on soil loss from five soils under natural rainfall

- We collected 84 storms with soil loss from five unit plots with different soils.
- The storms are divided into patterns as advanced, intermediate, delayed, and uniform.
- Storm pattern affects soil loss significantly with delayed pattern yielding most.
- Effect of storm pattern on soil loss is consistent within five Chinese zonal soils.

The effects of rainfall intensity on erosion processes for intra-storm variations are extensively investigated based on rainfall simulation experiments. However, there is a distinction between natural and simulated rainfall. Soil loss data from a total of 84 erosive storms from 2006 through 2013 with hyetograph rainfall data were collected from five plots located in Beijing, each with soils from different Chinese zones, to investigate the effects of erosive storm patterns on soil loss under natural rainfall conditions and to detect if this influence was consistent among the five soil types. The storms were divided into four patterns according to the period of the most concentrated rainfall, including advanced, intermediate, delayed, and uniform patterns. The results indicate that 1) the prevalent storm pattern in Beijing is advanced, which accounts for 43% of storms and contributes the most to total soil loss (approximately 55 to 68%). 2) Storm pattern significantly affected soil loss, with the delayed pattern yielding more soil loss than the other three patterns after taking into account EI30. 3) The effect of storm patterns on soil loss was consistent between the five soil types. 4) Coefficients for quantifying the effects of storm pattern on soil loss were not significantly different among these five soils, but exhibited significant differences among the storm patterns, e.g., the coefficient for the delayed pattern was 2.07, for the uniform pattern was 0.42, and for the advanced and the intermediate patterns were approximately 1. Rainfall erosivity generated by an erosive storm with a known pattern could be estimated by multiplying the corresponding adjustment coefficient, which would further improve the accuracy of soil erosion prediction, especially for storm level prediction.