figures used in the manuscripts.

Scripts to process the meteorological datasets to match the resolution of land surface model. As the land surface model used in this study uses its original encoding style (gtool format), decoding script is also included to convert to the plain-binary format for the hydrodyamic model.

Scripts to analyze and visualize the outputs. Mainly coded by Python.

The other datasets including model outputs and observation data are archived and publically accessible at (http://hydro.iis.u-tokyo.ac.jp/~yuta/EnsembleApproachTowardFlashFloodForecasting.tar.gz). If the file was too big to download or expand to your environment, the author would happily divide the required data from the archived file and provide them.

Any inquiry about this repository: please contact yuta@rainbow.iis.u-tokyo.ac.jp