Joint Frequency Distributions for Use in Erosion Research
1986
Istok, J. D. | Boersma, L.
Highest rates of soil erosion are often associated with specific climatic and soil conditions, e.g., rainfall on a frozen soil surface or on a saturated soil. Development of methods for computing the frequency of occurrence of these conditions using historical precipitation and air temperature data is reported. The methods are based on dividing hourly precipitation and air temperature data into precipitation events, freezing events, and event clusters. Precipitation events are defined as “wet” hours separated by “dry” hours. Freezing events are defined as periods of consecutive “cold” hours (air temperatures <0°C) separated by periods of consecutive “warm” hours (air temperatures >0°C). Event clusters are defined as over-lapping precipitation and freezing events. Sets of characteristics are defined for each type of event and for event clusters, e.g., duration and amount of precipitation for precipitation events, and freeze index and minimum temperature for freezing events. Joint frequency distributions for event and cluster characteristics are obtained, and examples show how to use these to predict frequency of occurrence of climatic conditions known to result in high rates of soil erosion. Application of joint frequency distributions to the Soil Conservation Service's curve number method for predicting watershed runoff is discussed, and it is shown how joint frequency distributions for precipitation amount and freeze index can be combined with the results of discriminant analysis for the occurrence of soil frost to compute the frequency of occurrence of rainfall on a frozen soil surface.
显示更多 [+] 显示较少 [-]