Combining geospatial data and numerical models to map and differentiate flooding extents caused by two tropical storms in the Philippines

By Meriam Makinano-Santillan, Jojene R. Santillan


In the year 2014, heavy rains associated with tropical storms Lingling (known as Agaton in the Philippines; January 10-20, 2014) and Jangmi (locally known as Seniang; December 27, 2014 – January 1, 2014) triggered massive flooding and caused fatalities in many localities, particularly in Caraga Region in northeastern Mindanao. Based on ground data, the 10-day Agaton event brought a total rainfall of 922 mm, of which 221 mm where recorded on January 19 alone, a day before it dissipated to the southeast of the Philippines. On the other hand, the 6-day Seniang event brought a total rainfall of 356 mm, of which 259 mm where recorded on December 29 alone. To better understand and differentiate the impacts of heavy rains brought by these tropical storms to the extent of flooding, we reconstructed the two flooding events by combining geospatial data from remote sensing and field surveys with numerical modelling. We focused on the Cabadabaran River Basin (CBR) and the nearby Pandanon River and Caasinan River Watersheds in Agusan del Norte, Caraga Region as our case study area. First, we developed a HEC HMS-based hydrological model of the river basin using a 10-m Synthetic Aperture Radar (SAR) Digital Elevation Model (DEM) for sub-basin delineations, land-cover maps from Landsat 8 OLI images for model parameterization, and rainfall and discharge datasets for model simulation and validation. The purpose of the HEC HMS model was to determine the volume of water coming from the various sub-basins that drains into the floodplains during the storms. The discharge hydrographs were then used as inputs into a 2-dimensional flood model to simulate the movement of flood water along the rivers and in the floodplains and to map the areas that are flooded. The 2D model was developed using a 1-m resolution LiDAR-derived Digital Terrain Model (DTM) and Landsat-derived landcover maps as its major parameters. From the numerical model simulations and output flood maps, we found that the Agaton event produced more discharge and caused wider extent of flooding than the Seniang event. This result is consistent with the fact that rainfall during Agaton was greater in volume than during Seniang. More areas were also in low and medium flood hazard levels (0.1 – 1.5 m depth) during Agaton. However, areas in high hazard levels (>1.5 m depth) appeared to be similar in both events. The results of this study showed the importance of combining geospatial data and techniques with numerical models to reconstruct and understand past flooding events. The flood simulations and maps derived from this study can be useful not only in flood hazard mapping of the project area, but also as visual aids to help people understand the differences of the impacts of different tropical storms in the occurrence of flooding.

KEYWORDS: geospatial data, remote sensing, numerical modeling, flood, tropical storms, Philippines

Publication Details/Metadata:


Authors: Jojene R. Santillan, Meriam Makinano-Santillan

Publication Date: 2015

Conference Name/Journal/Book Title: 13th Southeast Asian Survey Congress (SEASC 2015): Expanding the Geospatial Future

Publisher: ASEAN Federation of Land Surveying and Geomatics(ASEAN FLAG), Singapore Institute of Surveyors and Valuers (SISV)

Link to Full Text: View/Download