|
|
|
|
| LEADER |
01879 am a22001693u 4500 |
| 001 |
70830 |
| 042 |
|
|
|a dc
|
| 100 |
1 |
0 |
|a Hadipour, Sahar
|e author
|
| 700 |
1 |
0 |
|a Harun, Sobri
|e author
|
| 700 |
1 |
0 |
|a Arefnia, Ali
|e author
|
| 700 |
1 |
0 |
|a Alamgir, Mahiuddin
|e author
|
| 245 |
0 |
0 |
|a Transfer function models for statistical downscaling of monthly precipitation
|
| 260 |
|
|
|b Penerbit UTM,
|c 2016.
|
| 856 |
|
|
|z Get fulltext
|u http://eprints.utm.my/id/eprint/70830/1/SaharHadipour2016_Transferfunctionmodelsforstatistical.pdf
|
| 520 |
|
|
|a Three transfer function based statistical downscaling namely, linear regression model (LM), generalized linear model (GLM), generalized additive model (GAM) have been developed to assess their performance in downscaling monthly rainfall. Previous studies reported that performance of downscaling model depends on climate region and characteristics of climatic variable being downscaled. This has motivated to assess the performance of these three statistical downscaling models to identify most suitable model for downscaling monthly rainfall in the East coast of Peninsular Malaysia. Assessment of model performance using standard statistical measures revealed that LM model performs best in downscaling monthly precipitation in the study area. The Nash-Sutcliffe efficiency (NSE) for LM was found always greater than 0.9 and 0.7 with predictor set selected using stepwise multiple regression method during model calibration and validation, respectively. The finding opposes the general conception of better performance of non-linear models compared to linear models in downscaling rainfall. The near normal distribution of monthly rainfall in the tropical region has made the LM model much stronger compared to other models which assume that distribution of dependent variable is not normal
|
| 546 |
|
|
|a en
|
| 650 |
0 |
4 |
|a TA Engineering (General). Civil engineering (General)
|
