INGENIEUR
Figure 2: Simulated maximum, minimum and mean temperature from year 1980 â 2069 in
the Cameron Highlands
especially when the data set contains outliers.
An outlier markedly affects the mean while the
median is not influenced by them (Rumsey, 2011).
The median is a more appropriate measure of
central tendency when there are several outliers.
In order to minimise unnecessary errors, both
mean and median were taken as the measures of
central tendency in this study.
Results from Regional Climate Model were
analysed before bias correction of the data
was done to remove common projection model
errors. Past actual records of daily maximum and
minimum temperature and daily rainfall amount
data from MMD were taken into account as the
basis for analysis. Bias correction factors were
derived for the period 1983â2014 and applied to
1980â2069. The purpose was to remove common
systematic model errors. The bias correction
applied in this study is based on that proposed by
Terink et al (2010) for a Rhine basin study.
The bias correction is applied for precipitation
and temperature only. The corrected precipitation
P' was obtained as:
6
38
VOL
2017
VOL 70
55 APRIL-JUNE
JUNE 2013
Where:
= corrected precipitation
= mean unc orrec ted simulated
precipitation
= mean observed precipitation
= uncorrected simulated precipitation
Each daily temperature value T is transformed
to a corrected Tâ using:
Where:
= corrected temperature
= mean observed temperature
= standard deviation of observed
temperature
= standard deviation of uncorrected
simulated temperature
= uncorrected simulated temperature
= mean uncorrected simulated
temperature