The Correlation and Weight of Topography Also Rainfall Factors using Principal Component Analysis in Flood Analysis Nanda Khoirunisa (a*,b), Nurul Asyikin (a), Qori Fajar Hermawan (a,c), Zetsaona Sihotang (a,b), and Muhammad Riza (a,b)
a) Geophysics Study Program, Faculty of Mathematics and Natural Sciences, Mulawarman University
Jalan Barong Tongkok, Samarinda City, East Kalimantan 75242, Indonesia
*nandakhoirunisa[at]fmipa.unmul.ac.id
b) Oceanography Laboratory, Faculty of Mathematics and Natural Sciences, Mulawarman University
Jalan Barong Tongkok, Samarinda City, East Kalimantan 75242, Indonesia
c) Geophysics Laboratory, Faculty of Mathematics and Natural Sciences, Mulawarman University
Jalan Barong Tongkok, Samarinda City, East Kalimantan 75242, Indonesia
Abstract
The earth is vulnerable from climate change particularly sea-level rise and flood-related to extreme rainfall, yet most of the world^s coastal areas are growing faster than the inland population. The precipitation as a flood causatives factor mostly is the main influence of flood occurrence. The topography of an area also contributes to potential flood hazards in the nearest future. The digital elevation model and derivatives are elevations, slope degree, and aspect along with LANDSAT 8 derivatives are the normalized difference vegetation index (NDVI) along with the normalized difference vegetation index (NDWI). The investigation of flood disasters resulting from extreme rainfall and environmental factors regional macro information, in particular, utilize big data analysis with machine learning. The data input to principal component analysis (PCA) to determine the factor contribute to flooding hazard. The method is conducted to reduce the dimensions of data by working with only those indexes whose eigenvalues from PCA are greater than 1. The correlation value between -0.98 to 0.42 with the highest correlation is the elevation and the rainfall factors. In contrast, NDVI and NDWI have a strong negative correlation value. The weighted result shows the main drivers among the factor is rainfall, elevation, NDVI, aspect, slope, and NDWI with value 0.32, 0.29, 0.24, 0.20, 0,14, and -0.21, respectively. Further analysis reveals that the NDWI may not necessary and become the redundant factor for this group analysis.
Keywords: topography, rainfall, flood disaster, weight analysis, and principal component analysis
