Dataset and visualization scripts for DA-all (data assimilation using surface inundation depth and sewer water level ) experiments

This repository provides the dataset and visualization scripts used to generate the results for the DA-all experiments presented in the manuscript: Probabilistic urban flood forecasting with multivariate assimilation of sewer and surface observations The DA-all configuration applies data assimilation using both surface inundation depth and sewer water-level observations. The uploaded files include: h12_da_plot.tar.gz : Dataset containing model inputs and outputs h12_da_plot.ipynb : Jupyter Notebook for generating the visualizations in Fig. 4 of the manuscript ------------------------------------------------------------------------------------------------------------------------------------------ The dataset focuses on two synthetic observation scenarios (Case 1 and Case 4) and contains both pre-assimilation (NoDA) and post-assimilation (DA) outputs: git_mpi_urban_flood_202207_m_w_seed_noda_ens100 : Open loop results (NoDA) using 100 ensemble members. git_mpi_urban_flood_202207_m_w_seed_da_obs3_case1_idp : DA-all results derived using synthetic obs. case 1. git_mpi_urban_flood_202207_m_w_seed_da_obs3_case4_idp : DA-all results derived using synthetic obs. case 4. These Jupyter Notebook file contain the spatial distributions of surface inundation depth and sewer water-level changes before and after assimilation, corresponding to the fields visualized in Figure 4. The folder /h12_da_plot/git_mpi_urban_flood_202207_m_w_seed_noda_ens100/inputdata contains the model’s input files used consistently across all simulations: manhole.dat : Manhole data conduit.dat : Sewer conduit data bs_n_rn_nakahama.dat : Surface elevation data rain.dat : Rainfall forcing pump.dat : Pump data Additional files – Mesh generation data and h–A/R relationship tables for the sewer network