In the study of soil erosion , rain factor is one of the main agent. It is a necessity
to consider the matching of spatial and temporal scale when carrying out the study of
Soil Erosion Evaluation. It is generally thought a period of ten days to a month is the
temporal scale that matches the regional scale. However, using the temporal scale of a
period of ten days to a month to calculate gave rise to the generalization of rainfall
data and then exaggerate the error of erosion evaluation result. To solve one of the
contradiction, it is necessary to discuss monthly rainfall disaggregation method which
is suitable for regional soil erosion evaluation. Monthly rainfall disaggregation means
exploding and producing a group of rainfall intensity numerical sequence that enable
it to reflect true rainfall intensity and have the statistical characteristics similar to the
true rainfall. The base of this method is using theoretical distribution function to
describe the distribution characteristics of rainfall intensity numerical array. This
essay discussed the monthly rainfall disaggregation method based on the theoretical
distribution function appropriate for rainfall characteristics of Northern Shaanxi Loess
Plateau. The main conclusions are as follows :
(1)Three theoretical distribution functions of WBL、LOGN and GAMA are able
to represent monthly rainfall intensity distribution better with the use of historical
rainfall intensity data and parameter matching of maximum likelihood
method .Theoretical distribution have a low parameter estimation of the rainfall
intensity peak and the range of underestimation is 15% or so. (2)The representation ability of WBL distribution to the rainfall intensity
numerical sequence approaches 90%, while to the rainfall intensity peak is about 70%.
Thus is visible that this distribution function is suitable for describing monthly rainfall
intensity characteristics in the study of soil erosion on the Northern Shaanxi Loess
Plateau.
(3)Monthly rainfall disaggregation method can better realize the transformation
from monthly total rainfall to monthly rainfall intensity sequence. The transformation
result is close to statistical characteristics of true rainfall intensity. The difference
between the rainfall intensity sequence after rainfall disaggregation of most month
and statistical characteristics of true sequence is within 30%.
(4)Monthly rainfall duration (hourage) is the important parameter in the process
of rainfall disaggregation. This essay adopt monthly rainfall duration-rain capacity
relation to imitate and calculate rainfall duration ,which is evident in this study area.
Furthermore, the coefficient of determination R2 of them is about 0.7.As a result ,this
method suits for this area.
(5)The error in rainfall duration and monthly average rainfall intensity of most
month(about 70%)is between 20% and 30% when using theoretical distribution
function to conduct rainfall intensity disaggregation and acquired the estimated
monthly rainfall intensity numerical sequence.The average value of true rainfall
intensity peak is 8.8mm/h or so ,while of the estimated array is about 8.2mm/h. The
total tends to underestimation but the error isn't to a great extent. Besides, the
distribution trend of estimated rainfall intensity numerical sequence is more
concentrated.
Key Words：Regional Soil Erosion Evaluation; Rainfall Disaggregation; Theoretical
Distribution Function; Rainfall Intensity Numerical Distribution Characteristics
修改评论