Modelling of Sunshine Duration for Peninsular Malaysia
Knowledge of solar radiation at any area is important for designing solar energy conversion systems, and is also useful for ecophysiological studies. In the present study, we have analysed the daily sunshine duration records for Peninsular Malaysia and apply them in solar radiation modelling. Fi...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English English |
| 出版: |
1997
|
| 主題: | |
| 在線閱讀: | http://psasir.upm.edu.my/id/eprint/8623/1/FSAS_1997_8_A.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Knowledge of solar radiation at any area is important for designing
solar energy conversion systems, and is also useful for ecophysiological
studies. In the present study, we have analysed the daily sunshine duration
records for Peninsular Malaysia and apply them in solar radiation modelling.
Firstly, we fitted Angstrom type equation to solar radiation and
sunshine duration, to deter mine the relation between them. We have found
that the seasonal variations of the regression parameters, are different for
even two nearby locations. Variation of solar radiation within a small area
are also noticable. Secondly, we have studied on the probability distribution nature of
daily relative sunshine duration. Daily relative sunshine duration data are
fitted to three models. Parameters of the models are estimated only from the
monthly mean. Kolmogorov-Smirnov test is applied to determine the
goodness of fit. We have found that the four stations can model led by two
distributions. Data are also fitted to well known beta distribution model. The
effect of variance is also presented.
Finally, time series analysis of daily relative sunshine duration data
are presented in chapter 4. Using gaussian mapping technique, nonstationary
and non-normal distribution nature of the sunshine can be
transformed into stationary and normal distribution. The autocorrelation
function and partial autocorrelation function show the characteristics of
autoregressive process. Results from the Box-Pierce and Ljung-Box
statistics indicate that the first order autoregressive model is not suitable,
while the second order autoregressive model gives satisfactory results. The
relative sunshine duration of any day is dependent on the previous two
days. |
|---|