The amount of freely available satellite data is growing rapidly as a result of Earth observation programmes, such as Copernicus, an initiative of the European Space Agency. Analysing these huge amounts of geospatial data and extracting useful information is an ongoing pursuit. This paper presents an alternative method for flood detection based on the description of spatio-temporal dynamics in satellite image time series (SITS). Since synthetic aperture radar (SAR) satellite data has the capability of capturing images day and night, irrespective of weather conditions, it is the preferred tool for flood mapping from space. An object-based approach can limit the necessary computer power and computation time, while a graph-based approach allows for a comprehensible interpretation of dynamics. This method proves to be a useful tool to gain insight in a flood event. Graph representation helps to identify and locate entities within the study site and describe their evolution throughout the time series.

Insights into flood dynamics, rather than solely flood extent, are critical for effective flood disaster management, in particular in the context of emergency relief and damage assessment. Although flood dynamics provide insight in the spatio-temporal behaviour of a flood event, to date operational visualization tools are scarce or even non-existent. In this letter, we distil a flood dynamics map from a radar satellite image time series (SITS). For this, we have upscaled and refined an existing design that was originally developed on a small area, describing flood dynamics using an object-based approach and a graph-based representation. Two case studies are used to demonstrate the operational value of this method by visualizing flood dynamics which are not visible on regular flood extent maps. Delineated water bodies are grouped into graphs according to their spatial overlap on consecutive timesteps. Differences in area and backscatter are used to quantify the amount of variation, resulting in a global variation map and a temporal profile for each water body, visually describing the evolution of the backscatter and number of polygons that make up the water body. The process of upscaling led us to applying a different water delineation approach, a different way of ensuring the minimal mapping unit and an increased code efficiency. The framework delivers a new way of visualizing floods, which is straightforward and efficient. Produced global variation maps can be applied in a context of data assimilation and disaster impact management.

Describing spatio-temporal dynamics of a flood using an object-based approach with a graph-based representation proved useful for analysis of small-scale flood dynamics in Belgium (about 100 km²) (Debusscher, et al., 2019).  Starting from pre-processed Sentinel-1 SAR imagery, the water bodies are delineated in each timestep (using a thresholding algorithm), after which all water-polygons are grouped into graphs according to their spatial overlap on consecutive timesteps.  Change in (water)area and backscatter are used to quantify the amount of variation.  Products of this tool are a global variation map covering the whole study are, and a temporal profile for each waterbody, visually describing the evolution of the backscatter and number of polygons that make up the waterbody.  After establishing this proof of concept in a small region (flood of June 2016 in Schulensbroek, a nature reserve in north-eastern Belgium), this approach is applied on floods covering larger areas (about 10000 km²).  Two cases are studied, the Mozambique flood of March 2019 (near Beira) and the India flood of September 2019 (near Patna).  The process of upscaling leads to solving issues regarding the minimal mapping unit, adding extra pre-processing in order to simplify polygons (morphological operators), increasing code efficiency (mainly regarding for-loops).In the absence of ground truth, produced flood maps are compared to existing flood extent maps (from Disaster Charter (unitary) and Hasard (LIST)) in order to estimate accuracy.ReferencesDebusscher Bos and Van Coillie Frieke Object-Based Flood Analysis Using a Graph-Based Representation, Remote Sensing. - 2019. - p. p. 1883.