TY  - JOUR
T1  - Impact of Potential Evapotranspiration Method on Sensitivity and
Uncertainty in Streamflow Analysis for Kelantan River Basin
AU - Husain, Mohd Khardzir AU - Hayder, Gasim AU - Sidek, Lariyah Mohd AU - Ahmed, Ali Najah 
JO  - Journal of Engineering and Applied Sciences
VL  - 15
IS  - 5
SP  - 1228
EP  - 1236
PY  - 2020
DA  - 2001/08/19
SN  - 1816-949x
DO  - jeasci.2020.1228.1236
UR  - https://makhillpublications.co/view-article.php?doi=jeasci.2020.1228.1236
KW  - SWAT-CUP
KW  -SUFI-2
KW  -potential evapotranspiration
KW  -uncertainty
KW  -sensitive parameter
KW  -streamflow
AB  - Range of sensitive parameter value associated with uncertainty should be performed in the
calibration and validation process to establish more accurate watershed hydrological models. Evapotranspiration
is one of the most dominant in a hydrological cycle other than surface runoff and subsurface processes. This
research aims at an in-depth understanding of the SWAT-CUP ability on analyzing the impact of Potential
Evapotranspiration (PET) methods on sensitive parameter and uncertainty streamflow simulations of the
Kelantan river basin. The hydrological model was developed using the SWAT with three option of PET
methods: Penman-Monteith (P-M), Priestley-Taylor (P-T) and Hargreaves (HG). The SWAT-CUP with SUFI-2
optimization algorithm and Nash-Sutcliffe as objective function were used to evaluate the model simulation
compared to the streamflow discharge from years 1985-2000 for calibration and 2001-2016 for validation
periods. The best value of NSE, R2 and PBIAS, indicated no significant difference and the model achieved very
good performance during calibration and good performance during validation. The 95PPU plot and statistics
value, p-factor yielded acceptable outcomes during calibration by bracketing of the observed streamflow data
with 82, 74 and 75% for HG, P-T and P-M, respectively. However, the p-factor was achieved only 46, 40 and
44%, respectively during the validation period. The calibration strength of the r-factor was reached with
HG (0.75, 0.67), P-T (0.92, 0.84) and P-M (0.88, 0.81) during calibration and validation. The uncertainty
analysis showed that P-T is better performer during both the calibration and validation. Overall the SWAT
Model was considered can give good performance with the built-in PET methods options.
ER  -