Resilience Scanner

Integrated Flood Forecasting and Modeling System for River Basins

Solution Overview

A sophisticated flood forecasting and modeling system was developed for the Msimbazi River Basin to address recurring urban flooding challenges.[1][2][3] This integrated approach combines hydrological modeling, hydraulic simulation, and real-time forecasting capabilities to predict flood events and assess their impacts on vulnerable communities.[3][5] The system represents a collaborative effort to create more accurate flood hazard mapping and enable timely early warnings for areas experiencing frequent severe flooding.[5][7]

Technical Components

The technical architecture integrates multiple specialized modeling platforms and data sources to deliver comprehensive flood forecasting capabilities. SOBEK models river channels in 1D and inundation areas in 2D, with the 1D and 2D areas fully coupled.[2] The 1D component propagates floods in river channels while the 2D component models the inundation of land surface areas in terms of 2D flood propagation.[2] The SOBEK hydraulic flood model for the Msimbazi River uses a 20m resolution grid suitable for flood modeling, with LiDAR DTM 2019 data processed and resampled to this resolution.[2] A cascaded model composed of wflow sbm hydrological and SOBEK 1D2D hydraulic model was set up and calibrated for the Msimbazi River within the Tanzanian Urban Resilience Program.[2] The wflow sbm model was used to assess the catchment area of the Msimbazi river and the river discharges.[2] The Delft-FEWS system includes data feeds from observed hydro-meteorological data from TAHMO and forecasted meteorological precipitation from the NOAA Global Forecast System (GFS), a hydrological wflow model, a 1D2D urban flood model in SOBEK, and a Delft-FIAT impact assessment model.[3] The hydrodynamic model included detailed schematization of a community mapped drainage network and compared simulated flood depth with citizen's reports on flood depth for models run with and without the community mapped drainage network.[1] Available baseline data for model development included rainfall gauge data from 5 stations covering the period 1985-2018 with daily temporal resolution, though some gaps exist in the data.[4] Discharge data from 2 gauges covering the period 1985-2018 with daily temporal resolution is available, though some gaps exist in the data.[4] Water level data from 2 gauges with 15-minute temporal resolution is available for 2018, representing a short period of record.[4] A 30-meter resolution Digital Elevation Model from the 2000 Shuttle Radar Topography Mission is available, though the 2000 DEM may not accurately represent the current topography, especially in areas that have undergone significant development, and the Pugu Hills are not well represented.[4] Land use data from 2016 based on RapidEye imagery is available, though significant development since 2016 may have changed land use patterns.[4]

Implementation Details

The hydrological and hydrodynamic model of the Msimbazi Basin was developed for the Tanzania Urban Resilience Program Dar es Salaam Metropolitan Development Project.[4] Royal HaskoningDHV in association with CDR International and Hydro-Exuberance prepared the Inception Report for the Msimbazi Basin model development in September 2019.[4] The German International Cooperation Agency (GIZ) worked with OpenMap Development Tanzania (OMDTZ) and Dutch knowledge institute Deltares to collect cross-section surveys, extending an existing flood model for the Lower Msimbazi to include a much larger upstream portion of the river.[5] In collaboration with Deltares and with funding from GIZ, OMDTZ collected accurate and precise river cross-section surveys using an innovative, low-cost technique developed by the Resilience Academy and OMDTZ.[5] After the first field exercise to refine methods and safety precautions, OMDTZ was able to scale up and rapidly collect over 80 cross-sections of the river.[5] More than 300 students and locals were trained to use drones, GPS, and mobile apps to collect data for the Msimbazi Basin Development Project.[6] A new survey of the Msimbazi river cross-sections was required as the 2010 data was limited to a few sections and was not sufficient for developing a detailed hydrodynamic model.[4] The framework for urban flood modeling with community mapped data was demonstrated and validated on a case study in Dar es Salaam, Tanzania.[1] State-of-the-art urban flood models combine detailed geometric data on sewer pipes and drainage channels with digital terrain data to model flood events, combining both overland flow and channel flow through channels and pipes.[1]

Benefits and Impacts

Delft-FEWS runs operationally at Deltares to provide a forecast every hour for the following 8 hours on river depth, flooded area, number of buildings affected per neighborhood (ward), flooded area of the bus depot, and flooded bus stops.[3] A lead time of 8 hours was chosen because it takes water approximately 8 hours to flow from the very upstream parts of the Msimbazi River into the city of Dar es Salaam, which is enough time for stakeholders to act on forecasted floods while avoiding many false alarms and misses.[3] The OMDTZ survey of river cross-sections and hydraulic structures along the Msimbazi River supported the creation of a comprehensive hydraulic model of the area that has led to the development of a more realistic flood hazard map of the most vulnerable areas.[5] Community Water Watch (CWW) is a community-designed and -operated flood early warning service for Dar es Salaam, Tanzania that uses hydrological and hydrodynamic models built upon locally-collected map data.[3] GFDRR supported real estate assessments showing that flood protection in the Msimbazi Basin could unlock up to $900 million in investment.[6]

Climate Adaptation Relevance

Each year, Msimbazi flooding leads to fatalities and destroys critical infrastructure.[5] Over the past decade, the middle and lower part of the Msimbazi Basin has been experiencing frequent, severe flooding at least twice a year during every rainy season, and in 2019 alone, nine major flood events took place.[7] The integrated modeling system addresses extreme precipitation and flooding hazards by providing hourly forecasts with sufficient lead time for emergency response and community preparedness actions.[3]

[1]: Citation 1 [2]: Citation 2 [3]: Citation 3 [4]: Citation 4 [5]: Citation 5 [6]: Citation 6 [7]: Citation 7

Sources