Land cover change and the use of SoilGrids data for assessing flash flood in Marikina river basin, The Philippines : abastract

Severe flooding causes damage to properties, disruption of economic activities and loss of lives. In the Philippines flooding is of particular concern because it can affect heavily populated areas like Manila. Excessive rain is often the major cause but anthropogenic activities such as deforestation, land use changes, and urbanization also affect flood risk.
Event based models help in understanding the interactions of falling rain on vegetation cover and soil in generating surface runoff and stream discharge from a catchment. If sufficient good quality data is available it is possible to generate flood hazard maps and scenarios of extreme rain events. But data availability is often crucial problem in many places. Meteorological stations generally collect daily rainfall data which is not suitable to simulate floods. Similarly, land use maps can be outdated and soil maps may not be available. In the absence of laboratory analysis data on soil, parameters required to run the model are often generated from soil texture using pedo-transfer functions. Recently 250 m soil grid data, based on global covariates and globally fitted models, is available (www.soilgrids.org) which eases the data scarcity problem. However, it is based on a self-learning statistical method and its quality depends on the original data density. In this research we study the sensitivity of using SoilGrid and other sources of soil related information for flashflood hazard in the Marikina River Basin. Also the influence of land cover change is studied. OpenLisem is applied to generate flood scenarios.
The results shows that SoilGrid is a useful data source solving data scarcity problems. However, its use should be supported by some soil sampling and lab analysis. The results also show that vegetation cover change has an insignificant effect during convective or extreme conditions. In the downstream part, urbanization has an effect on flood extent, volume and flood duration.