A detailed technical description of the three state-of-the-art SAR-based GFM flood mapping algorithms - which were developed by the GFM consortium members LIST, DLR and TU Wien - is provided in the GFM Product Definition Document (PDD), together with examples of their application in an operational context. This section of the PUM summarises the key technical points of the three individual GFM flood mapping algorithms and provides links to the more detailed information in the PDD.
All three of the GFM flood mapping algorithms - namely Algorithm 1 (LIST), Algorithm 2 (DLR) and Algorithm 3 (TU Wien) - make use of historical time-series of Sentinel-1 SAR intensity data and use topography-derived indices to refine the initial classification of water bodies. However, differences appear in the ways historical time series of intensity data are finally used to parameterize the retrieval algorithms and the way ancillary data such as topography data are used in the production system. Other differences relate to the inclusion of a region-growing step or not, the scale at which the thresholds are determined and applied to each pixel’s backscatter value and other nuances in the way the retrieval algorithms are setup.
Each GFM flood mapping algorithm takes as input Sentinel-1 (S-1) image data, and generates an “S-1 Observed Flood Extent” output layer, which is further refined using the “Exclusion Mask” (to mask out areas that cannot be classified, due to local conditions), and the “S-1 Reference Water Mask” (showing permanent and seasonal water bodies).
An ensemble-based approach is then used to combine the observed flood extent maps that are generated by the three GFM flood mapping algorithms, into single “consensus maps” representing the final GFM product output layer “S-1 Observed Flood Extent”.
Similar to the GFM ensemble flood mapping algorithm, which is used to generate the output layer “S-1 Observed Flood Extent”, a GFM ensemble water mapping algorithm is used to generate the output layer “S-1 Reference Water Mask” (showing permanent and seasonal water bodies), based on a “data cube” or time series of Sentinel-1 image data.
The (internal) availability of three separate maps for observed flood extent also provides a convenient means of identifying and addressing any shortcomings affecting a single GFM flood mapping algorithm in specific circumstances and / or parts of the world, due to many well-known factors like topography or environmental conditions. Accordingly, the implemented quality assurance procedures are able to investigate whether classification errors can be attributed to shortcomings of individual algorithms, or to limitations that are inherent in the Sentinel-1 SAR sensors, and their ability to capture the appearance (or absence) of surface water in particular situations.
The most relevant features of the three main GFM flood mapping algorithms are summarized and compared in Table 1 below. In the following sub-sections, more details are provided on the three main GFM flood mapping algorithms, and on both the GFM ensemble flood mapping algorithm, and the GFM ensemble water mapping algorithm, which are used, respectively, to produce the “S-1 Observed Flood Extent” and “S-1 Reference Water Mask” output layers.
| Algorithm |
ALGORITHM1 |
ALGORITHM2 |
ALGORITHM3 |
| Provider | LIST | DLR | TU Wien |
| Application domain | Water and flood extent mapping (pixel-based) | NRT Water and flood extent mapping | Pixel-based flood extent mapping |
| Input remote sensing data | Pair of SAR intensity images acquired from same orbit (any sensor) and model parameters derived from historical time series | Single-temporal SAR intensity data | Single SAR acquisition and model parameters derived from historical time series |
| Auxiliary data | HAND index map, exclusion layer, reference water layer, water and flood extent map computed at previous time step | HAND index exclusion map, reference water extent, DEM, optional: low backscatter exclusion mask based on S-1 time-series data | HAND index, exclusion mask, reference water map for generating the fresh flooded areas |
| Main characteristics | Scene-specific statistical modelling of backscatter distributions, systematic updating of water bodies maps using combination of change detection and region growing | Hierarchical automatic tile-based thresholding, fuzzy logic-based post classification and region growing | Classification based on backscatter probability distribution by exploiting the historical time series with consideration of backscatter seasonality. |
| Exploits time series of SAR data | Yes | No, the integration of a low backscatter exclusion mask based on S-1 time-series data (produced offline) can be integrated optionally | Yes, parametrisation through multi-year time series |
| Exploits textual information by region growing | Yes | Yes | No |
| Automation | High | High | High |
| Initialization | Statistical modelling of backscatter distributions attributed to water / no water and change / no change classes (per tile) | Hierarchical automatic tile-based thresholding using statistical modelling of class distributions | Generation of backscatter probability distribution from historical time series measurements |
| Post-classification steps | Masking of exclusion areas, distinction between water and flood extent using reference water layer | Masking of exclusion areas, distinction between water and flood extent using reference water layer | Noise reduction, Mask the exclusion areas, extraction fresh flood area compared with reference water map |
| Water probability mask generated | Yes (based on Bayesian inference) | Yes (based on fuzzy logic) | Yes (based on the Bayesian posterior probability) |
| Outstanding/differentiating features | Hierarchical split-based approach enabling re-calibration of parameters in NRT based on most recent pair of S-1 images | Fuzzy logic-based approach enabling a post classification and region growing taking advantage of topography-derived indices in addition to SAR backscatter | Exploiting per-pixel full Sentinel-1 signal history in data cube; enabling a very fast and scalable production of flood and water extent maps through pre-computed global parameters at high quality |
| Additional information | |||