Abstract:
This thesis applied the hydrological model SWAT to simulate the rainfall - runoff process 111
the Upper Ping River Basin which covers the area of about 23,370 sq.km. The purpose of
the simulation was to observe the influences of changes in hydrometeorologic conditions
and in land uses on the OCCUlTence of river runoff. It is expected that the results might be
used to predict the change in river runoff in the future when the land uses pattern change
and that the feasibility and adequacy of the present water usage plans and water
management scheme can be assessed.
In the study, statistical analyses on trends of changes in hydrometeorologic data were done.
The SWAT model was then set up for the river basin. The whole river basin which has
the Bhumibol Dam as the outlet was divided by SWAT into 11 sub - basins. Six subbasins
out of these 11 sub - basins have the streamflow gaging stations or the streamflow
data at their outlets. They are the P.l, P.14, P.20, P.2l streamflow gaging stations and the
two storage dams, Mae Ngad Dam and Mae Guang Dam.
Model Calibration was done by uS1l1g the hydrometeorologic data and land uses data of the
years 1990, 2001 and 2006. The computed river runoff values were compared to the
measured data at streamflow gaging stations P.l, P.14, P.20 and P.2l and at Mae Ngad
Dam and Mae Guang Dam. Justification of the model calibration was considered from the
efficiency coefficient Nash and Sutcliffe (NS) and the correlation coefficients (R\
Calibration result of the year 1990 yielded R2 between 0.61 - 0.89 and NS between 0.74-
0.94 while that of the year 2001 yielded R2 between 0.68 - 0.83 and NS between 0.82-
0.91. Thc model parameters used for calibration in 2001 were then verified with runoff data
2 in 2006. Thc verification results that showed R ranged between 0.71 0.93 and NS
between 0.81 0.95. The values of these two coefficients. as they fell within the acceptable
range. indicated that the values of model parameters as used in the calibration could
produce the values of river runoff closely similar to the observed ones.
The calibrated model was then used to simulate the occurrence of river runoff 111 Upper
Ping River Basin under three different land use patterns. They are those of 1990, 200 I and
2006. For each type of land use, simulation was done by using the past 20 years of
meteorologic data (1987-2006).
The simulated results showed that land uses pattern affects the river runoff volume but its
effect was not clear on the distribution pattern of monthly runoff volume. The annual
runoff volume of 2006 land uses was greater than the ones of 1990 and 2001 land uses.
This was in agreement with the trend analysis of the runoff data measured at streamflow
gaging station P.l and at Bhumibol Dam. However, due to the fact that three patterns of
land uses which were used in the study contained just minor differences in the percentages
of land occupied by each type of land use and that the trend of change could not be
clearly identified. the impact of land uses on the rainfall runoff process in the Upper Ping
River Basin still cannot be clearly concluded from this study.