Homogeneous climate divisions for Peninsular Malaysia
Classification of Peninsular Malaysia was delineated by integrating in-situ temperature elements data and Geographical Information System (GIS) raster data. The principal component (PC) analysis was applied to long-term mean monthly temperature elements data for monsoon seasons. The first three prin...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Published: |
Taylor & Francis
2011
|
| Subjects: | |
| Online Access: | http://eprints.um.edu.my/23032/ https://www.tandfonline.com/doi/abs/10.3166/ga.24.89-94 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Classification of Peninsular Malaysia was delineated by integrating in-situ temperature elements data and Geographical Information System (GIS) raster data. The principal component (PC) analysis was applied to long-term mean monthly temperature elements data for monsoon seasons. The first three principal components were chosen to be statistically significant accounted for 96.5% of the variability in the 27 variables. These three components are related to the mean monthly variation in minimum temperature during monsoon season (first PC) the mean monthly variation in maximum and the mean temperature in southwest monsoon (second PC) and the mean monthly variation in maximum temperature during northeast monsoon (third PC). Cluster analyes were applied to create clusters of meteorological stations of which six classes were formed. To determine cluster boundaries interpolation analysis was applied to generate GIS raster data of factor scores. The supervised classification analysis was then performed to the generated GIS factor data. The result of a maximum likelihood classification produced three clusters when summarized by districts. Final classification results of climate divisions show rational climate regionalization that reveals control on temperature. The use of factor score GIS raster data effectively assists the generation of meteorological station clusters grouped using only in-situ data. © 2011 Lavoisier SAS. All rights reserved. |
|---|