Regression models for the
k and
n values of the
L=
kQn-type pollutant runoff model (
L : specific load,
Q : specific discharge,
k and
n : constants) were developed. In the models, the independent variables are several watershed characteristics etc.
The relation between the
k value and several watershed characteristics was analyzed using the data on 14 rivers with 17 stations as well as the
n value. According to multi-regression analyses for the
k and
n values, it was found that (1) for the
k value a correlation coefficient of a regression model was large even if discharged load was an only independent variable and (2) for the
n value a correlation coefficient of a regression model was large when the
k value or discharged load which had a strong correlation with the
k value was included as an independent variable.
Regression models for estimating the
k and
n values were selected from the results and verified by observed data. According to the verification, the estimated
k and
n values ranged from half of the observed values to twice of those at most of the stations as well as the estimated load which was calculated by the estimated
k and
n values.
View full abstract