Application of Precipitation Downscaling Methods in Hydrological Model
Hao Zhenchun (1), Su Fenge (1)
(1) Water Resources Development & Utilization Lab., Hohai University
Due to the increase of time interval and decrease of precipitation intensity within a month, no monthly runoff appears in some gridded cells as the Xin'anjiang monthly hydrological model is applied to the Huaihe River Basin.
The best way to solve the above problem is to run the model at a daily time step with daily input. Unfortunately, in most cases, only monthly-observed average data are available. In this study, the Xin'anjiang model is run at a daily time step from monthly data, and the model output is at a monthly interval. Two methods of downscaling of monthly precipitation to daily resolution are considered and the model results are compared.
The first is a simple stochastic method, where the rainfall on rain-days is set to be equal to the average daily intensity in that month. So the magnitude of rainfall on rain-days can be expressed as the following formula:
P/Pdays (IWhere PP(I) is the magnitude of rainfall on rain-days; P is monthly precipitation; Pdays is the number of rain-days.
The second method stochastically generates the magnitude of rainfall on the rain-days from an exponential distribution:
e-x
Where x is the magnitude of rainfall on rain-days; and are coefficient .So the rainfall amount on rain-days can be expressed as the formula:
F-1(ui)Where F-1 is the inverse function of F; ui is uniform [0,1] random number; 1 and
1 are coefficient .
The number of rain-days in any month is estimated using an empirical relationship developed from rainfall data in the Huaihe River basin. The relationship can be quantitatively expressed as the following formula:
2*DN(J)/?*Arctg(A*P+B)
1,2,c,12); A and B are coefficients.
In this presentation, two methods of downscaling monthly precipitation to daily resolution are validated by running the Xin'anjiang model at a daily time step from the monthly data, and the model outputs are more accurate than the monthly hydrological model.
In this study, the probability of rain on any day is considered to be equal. So the generated daily precipitation is distributed at random through a month. In the future study, the generated daily precipitation should be distributed more reasonable based on the analysis of real data.
Submittal Information
Name :
Date :
Prof. Hao Zhenchun
27-May-01-13:07:06
Organization :
Theme :
Water Resources Development & Utilization Lab., Hohai University