This dataset was generated by the Remote Sensing Group of the TU Wien Department of Geodesy and Geoinformation (https://mrs.geo.tuwien.ac.at/), within a dedicated project by the Join Research Centre (JRC) of the European Commission. Open use is granted under the CC BY 4.0 license. End of summer 2022 Pakistan was hit by one of the most severe floods in decades. The event was covered by multiple satellite-based emergency services, including the Copernicus Emergency Management Service (CEMS) global flood mapping (GFM) component. As part of the project's consortium, the Technische Universität Wien (TU Wien) developed a dedicated flood mapping algorithm (Bauer-Marschallinger et al. 2022) using the Synthetic Aperture Radar (SAR) satellite Sentinel-1 as an input. The published dataset contains the results of the TU Wien algorithm for the time period August 10 to September 23, 2022 and the covered area is located in the southern part of Pakistan. Besides the binary flood maps, the dataset contains retrieved statistics aiming for presenting the impact of the event as seen from satellite data. With the publication of this dataset, we want to share timely results of our algorithm and support further studies about the event. It is planned to publish the dataset alongside of a dedicated paper in the journal for "Natural Hazards and Earth System Sciences". Within in this publication, the flood mapping results were evaluated based on the results of the CEMS rapid mapping component. Dataset Record Flood mapping results The flood mapping results (FLOOD-HM-MASKED.zip) are sampled at 20 m pixel spacing, georeferenced to the Equi7Grid and divided into tile of 300km extent ("T3"-tiles). The used folder structure splits up the single file per Equi7Grid tile and the used filenaming can be interpreted as follows:

VAR_TIME__POL_ORBIT_TILE_GRID_VERSION_SENSOR_CREATOR.tif Where:

VAR: Variable name ("FLOOD-HM-MASKED") TIME: Acquisition time of the used Sentinel-1 scene POL: Polarisation of the used Sentinel-1 scene ("VV") ORBIT: Orbit direction ("A" for ascending and "D" for descending) and relativ orbit number of the used Sentinel-1 scene TILE: Equi7Grid tile code GRID: Equi7Grid continent subgrid VERSION: Software version and run number SENSOR: Used sensor ("S1") CREATOR: Creator of the file ("TUWIEN") The values of each files can be interpreted like this:

0: no flood 1: flood 255: masked Flood statistics The dataset consists of two statistical layers: the flood frequency (flood_frequency.tif) and the time of the first flood detection (first_detection.tif). Both layers are available as merged file for the whole study area and georeferences in the WGS84 coordinate system. The flood frequency is known as the ratio of number of flood detection and number of valid observations of a pixel and is given in percentage in this case. It provides insights about the continuity and duration of a flood classification at a pixel level. For instance, the area which was flooded at least once or during the whole time period can be extracted. The time of the first flood detection is given as day-of-year (DOY) and indicates the day when the first flood detection was found for a specific pixel. This information can be used to get insights about the progress of the flood. Related Software

The flood mapping results are provided in the Equi7Grid, which is accessible by its dedicated open-source package: https://github.com/TUW-GEO/Equi7Grid  The folder structure and filenaming of the flood mapping results is defined and handled by: https://github.com/TUW-GEO/geopathfinder The metadata definition of the flood mapping results is collected in: https://github.com/TUW-GEO/medali The flood mapping results follow a dedicated folder and filenaming structure, which can be handled easily by: https://github.com/TUW-GEO/yeoda Acknowledgements This study was funded by TU Wien, with co-funding from the project "Provision of an Automated, Global, Satellite-based Flood Monitoring Product for the Copernicus Emergency Management Service" (GFM), Contract No. 939866-IPR-2020 for the European Commission's Joint Research Centre (EC-JRC). The computational results presented have been achieved using i.a. the Vienna Scientific Cluster (VSC). References Bauer-Marschallinger, B., Cao, S., Tupas, M. E., Roth, F., Navacchi, C., Melzer, T., Freeman, V., and Wagner, W.: Satellite-Based Flood Mapping through Bayesian Inference from a Sentinel-1 SAR Datacube, Remote Sensing, 14, 3673, 2022.