INGENIEUR
cumecs , the lowest in the past five to ten years . Average Recurrence Interval ( ARI ) ( Hydrology and Water Resources Division Department of Irrigation and Drainage Malaysia , 2005 ) .
( iii ) Land Surface Temperature ( LST ) LST distribution is strongly linked to NDVI distribution , where LST typically provides an opposite relation to NDVI . This means that with low NDVI ( less vegetation ), LST increases in temperature . From the values acquired from the analysis , it showed that the temperature during dry months are relatively higher compared to wet months as shown in Figure 5 . Increase in LST is strongly related to the increase of urbanisation . Due to this , it is considered natural to have low LST values in the earlier years . In this case , the lowest value occurred in the year 1988 , where there was less urbanisation and highest LST in more recent years . A sudden increase in the second point for both dry and wet months with a temperature of
26.2 0 C and 20.1 0 C might be due to replacement of natural land cover and cutting down of forests in the area for oil palm planting .
( iv ) Unsupervised Clustering Classification In total , for the 32-year observation period , a comparison was made between 1988 data and 2020 data . It was found that the proportion of vegetation with high NDVI values (> 0.6 ) had the largest increase with an increase of 106,200 square metre followed by sparse vegetation with high NDVI values (> 0.45 ). Sparse vegetation with low NDVI values (< 0.45 ) however , decreased by 148,00 square metres in area . Figure 6 shows the comparison in the vegetation change in the area of interest .
A confusion matrix was generated to analyse which areas in the oil palm land scheme had developed in land cover and which areas had stayed the same during the period of 32 years . Figure 7 shows the combined classification map
Figure 5 : LST trendline for dry and wet months from year 1988-2020 .
56 VOL 84 OCTOBER-DECEMBER 2020