Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Urban flood modelling : a GIS based approach in Lomma, Skåne region

Ghebremariam, Michael Araya LU (2017) In Student thesis series INES NGEM01 20171
Dept of Physical Geography and Ecosystem Science
Abstract
Urbanization triggers flooding because it replaces pervious land surfaces with impervious surfaces that have less capacity to infiltrate and store water into the ground. Detailed analysis and modelling of flooding in urban areas can be performed with GIS-based distributed hydrological models. However, the implementation of these models require high-level proficiency in GIS and hydrology. Thus, many municipalities hire consultants to do the task. In this study, a GIS-based model that can perform urban flood modelling is developed.

The process begins with pre-processing the original DEM so that it represents the urban terrain with its constructed streets and building that can change the direction of drainage path. This is
done through... (More)
Urbanization triggers flooding because it replaces pervious land surfaces with impervious surfaces that have less capacity to infiltrate and store water into the ground. Detailed analysis and modelling of flooding in urban areas can be performed with GIS-based distributed hydrological models. However, the implementation of these models require high-level proficiency in GIS and hydrology. Thus, many municipalities hire consultants to do the task. In this study, a GIS-based model that can perform urban flood modelling is developed.

The process begins with pre-processing the original DEM so that it represents the urban terrain with its constructed streets and building that can change the direction of drainage path. This is
done through integrating spatial data using Arc Hydro tools. Flow routing of water over the modified DEM is then calculated using two flow direction algorithms (1) the deterministic eight-node algorithm (D8) and (2) the Triangular Form Based Multiple Flow algorithm (TFM). D8 algorithm assumes that flow at a point follows only the steepest downhill slope to one of the eight possible directions. TFM algorithm on the other hand estimates flow distribution values proportionally to the slope gradient in each direction. The effect of street inlets flow interception is introduced in the analysis through preparing a weight raster. To develop the weight raster, two approaches are applied depending on which algorithm is used to assign flow direction over the terrain. In D8 method, flow accumulation at street inlet points is calculated from the sub-watershed areas delineated for each inlet point. Weight value assignment of cells within a sub-watershed is then calculated as a proportion of volume to be consumed by street
inlet to the total volume generated from the sub-watershed. In TFM method, flow accumulation at inlets is calculated by running TFM algorithm with input DEM having since cells at locations of street inlets. Weight value assignment of cells is done by reducing the flow which is intercepted by inlets from the flow accumulation value at sink cells. In addition, the capacity of street inlets to intercept flow is also estimated.

This study shows the capability of performing flooding in GIS environment. The results also show that the outcome of GIS-based urban flood modelling is different depending on which algorithm is used to calculate flow direction. Results of flow accumulation before the inclusion of street inlets interception effect in the analysis is 57,271 m3 and 45,028 m3 using D8 and TFM methods respectively. After street inlets interception effect is included in the analysis however, the results show that weighted flow accumulation is reduced to 33,316 m3 and 10,893 m3 using D8 and TFM methods respectively. In addition, 202 flooding incidents at sink cells are identified using D8 method, this number drops to 80 sink cells using TFM method. (Less)
Popular Abstract
Simplified GIS-based models for urban flood modelling

Urbanization replaces pervious surfaces by impervious ones having low capacity to infiltrate and store water into the ground. GIS-based distributed hydrological models such as PCSWMM and Mike Urban can perform a detailed analysis and modelling of flooding in urban
areas. However, the implementation of such models requires high-level proficiency in hydrology and GIS. In addition, this approach is generally beyond the municipality’s budget in time and cost. For this reason, municipalities often hire consultants to do this. In this study, simplified GIS-based models are developed to perform flooding in urban environment.

DEM can be used for watershed boundary delineation and... (More)
Simplified GIS-based models for urban flood modelling

Urbanization replaces pervious surfaces by impervious ones having low capacity to infiltrate and store water into the ground. GIS-based distributed hydrological models such as PCSWMM and Mike Urban can perform a detailed analysis and modelling of flooding in urban
areas. However, the implementation of such models requires high-level proficiency in hydrology and GIS. In addition, this approach is generally beyond the municipality’s budget in time and cost. For this reason, municipalities often hire consultants to do this. In this study, simplified GIS-based models are developed to perform flooding in urban environment.

DEM can be used for watershed boundary delineation and drainage pattern extraction in rural environment. However, in urban areas similar analysis is complicated due to the constructed terrain of streets and buildings which can change the direction of drainage path. Therefore, the original DEM has to be modified so that it represents the urban terrain more accurately. This
is done in the DEM pre-processing step through integration of spatial data such as street inlets, buildings and the DEM in Arc Hydro tools and ArcGIS. In the next step flow routing of water over the terrain is calculated using two flow direction algorithms namely the Deterministic eight-node (D8) algorithm and Triangular Form Based Multiple Flow Direction (TFM) algorithm. D8 algorithm assumes that in a 3 X 3 cells moving window flow at a point follows only the steepest downhill slope to one of the eight possible directions.
TFM algorithm on the other hand estimates flow distribution values proportionally to the slope gradient in each direction. Flow interception by the stormwater collection points (street inlets) is included in the analysis. To do this street inlets flow interception capacity is estimated and used to develop a weight raster. The developed weight raster is then used as input weight during flow accumulation calculation. To prepare a weight raster input using D8
flow direction algorithm, first sub-watershed areas contributing flow to each of the street inlets are delineated and flow accumulation generated by the sub-catchments is calculated. Weight value assignment of cells within a sub-watershed is then calculated as a proportion of volume to be consumed by street inlet to the total volume generated from the sub-watershed. In TFM method, first sink structures are created at locations of street inlet points. This DEM with sinks is then used to calculate flow direction and flow accumulation in MATLAB. Flow accumulation values at the sink cells are then extracted and used to prepare weight raster. Weight value assignment is done by reducing a value which is equal to flow interception capacity of street inlets from the extracted flow accumulation values at street inlets. The weight raster is used as input to run flow accumulation with weight using a different tool that can operate with negative value.

This study shows the capability of GIS to perform flooding in urban area. The results also show that GIS-based flood modelling produce different results depending on which algorithm is used to calculate flow direction, Flow accumulation before the inclusion of street inlets interception effect in the analysis is 57,271 m3 and 45,028 m3 using D8 and TFM methods respectively. After street inlets interception effect is included in the analysis however, the
results show that weighted flow accumulation of 33,316 m3 and 10,893 m3 using D8 and TFM methods respectively. In addition, 202 flooding incidents at sink cells are identified using the D8 method, this number drops to 80 sink cell using TFM method. (Less)
Please use this url to cite or link to this publication:
author
Ghebremariam, Michael Araya LU
supervisor
organization
course
NGEM01 20171
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Physical Geography and Ecosystem analysis, Digital Elevation Model, Single flow direction algorithm, Deterministic eight-node (D8) algorithm, Multiple flow direction algorithm, Triangular Form Based Multiple Flow algorithm (TFM)
publication/series
Student thesis series INES
report number
439
language
English
id
8929122
date added to LUP
2017-12-05 11:27:43
date last changed
2017-12-05 11:27:43
@misc{8929122,
  abstract     = {{Urbanization triggers flooding because it replaces pervious land surfaces with impervious surfaces that have less capacity to infiltrate and store water into the ground. Detailed analysis and modelling of flooding in urban areas can be performed with GIS-based distributed hydrological models. However, the implementation of these models require high-level proficiency in GIS and hydrology. Thus, many municipalities hire consultants to do the task. In this study, a GIS-based model that can perform urban flood modelling is developed.

The process begins with pre-processing the original DEM so that it represents the urban terrain with its constructed streets and building that can change the direction of drainage path. This is
done through integrating spatial data using Arc Hydro tools. Flow routing of water over the modified DEM is then calculated using two flow direction algorithms (1) the deterministic eight-node algorithm (D8) and (2) the Triangular Form Based Multiple Flow algorithm (TFM). D8 algorithm assumes that flow at a point follows only the steepest downhill slope to one of the eight possible directions. TFM algorithm on the other hand estimates flow distribution values proportionally to the slope gradient in each direction. The effect of street inlets flow interception is introduced in the analysis through preparing a weight raster. To develop the weight raster, two approaches are applied depending on which algorithm is used to assign flow direction over the terrain. In D8 method, flow accumulation at street inlet points is calculated from the sub-watershed areas delineated for each inlet point. Weight value assignment of cells within a sub-watershed is then calculated as a proportion of volume to be consumed by street
inlet to the total volume generated from the sub-watershed. In TFM method, flow accumulation at inlets is calculated by running TFM algorithm with input DEM having since cells at locations of street inlets. Weight value assignment of cells is done by reducing the flow which is intercepted by inlets from the flow accumulation value at sink cells. In addition, the capacity of street inlets to intercept flow is also estimated.

This study shows the capability of performing flooding in GIS environment. The results also show that the outcome of GIS-based urban flood modelling is different depending on which algorithm is used to calculate flow direction. Results of flow accumulation before the inclusion of street inlets interception effect in the analysis is 57,271 m3 and 45,028 m3 using D8 and TFM methods respectively. After street inlets interception effect is included in the analysis however, the results show that weighted flow accumulation is reduced to 33,316 m3 and 10,893 m3 using D8 and TFM methods respectively. In addition, 202 flooding incidents at sink cells are identified using D8 method, this number drops to 80 sink cells using TFM method.}},
  author       = {{Ghebremariam, Michael Araya}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Student thesis series INES}},
  title        = {{Urban flood modelling : a GIS based approach in Lomma, Skåne region}},
  year         = {{2017}},
}