Directions Magazine
Hello. Login | Register

Articles

All Articles | Post Comment

The Urban Drainage Model: SF DPW uses LiDAR DEM and a Custom Algorithm for Delineating Drainage Catchments and Hydrologic Modeling

Wednesday, June 15th 2011
Comments
Classified Ads:

Summary:

This article describes how the San Francisco Department of Public Works used a very high-resolution digital elevation model to study urban storm water hydraulics. The base data came from a 1-meter horizontal resolution LiDAR aerial survey of San Francisco that was acquired in June 2007. The FBI ordered the survey prior to that year’s baseball All-Star game for use in security analysis, and provided the data to the city for free after the event.

 

Ed. note: This article was original published in the Spring 2011 edition of The BAAMA Journal, the biannual publication of the Bay Area Automated Mapping Association (BAAMA).

One of the uses of a Digital Elevation Model (DEM) is to automatically generate catchment areas for use in hydraulic modeling. This type of analysis is often used to model rivers and the horizontal resolution in such instances tends to be coarse, typically 30 meters or greater. Last year, San Francisco Department of Public Works (SF DPW) applied similar techniques at a fine resolution. The results are currently being used to study pipe hydraulics for San Francisco’s sewer system.

A 1-meter horizontal resolution LiDAR (Light Detection and Ranging) aerial survey of San Francisco was acquired in June 2007. The FBI ordered the survey prior to that year’s baseball All-Star game for use in security analysis, and provided the data to the City for free after the event. When studying urban storm water hydraulics, especially in a very hilly area like San Francisco, a high-resolution base layer is an ideal source for creating accurate models. SF DPW used the survey to delineate drainage areas across the entire city at the level of individual drain inlets, which were then aggregated to larger basins.
   
When creating a small scale catchment delineation of a large area, traditionally, an experienced engineer may review map contours, the sewer network, property lines, and other data. Then that engineer may draw the catchment boundaries based on the information at hand and his or her own subjective experience. If the engineer does not have time to review and delineate the entire area of interest, a team may finish the work. Since each team member would do delineation slightly differently and introduce unintended biases, this process introduces some uncertainty. A major benefit of delineating drainage boundaries using a DEM or LiDAR is the process can be replicated.  

To automate the drainage catchment delineation process, the authors created a computer script in ArcGIS ModelBuilder called the “Urban Drainage” model. Several GIS layers from SF DPW’s Sewer Information System are used as inputs to the script including manhole, drain inlet, and pipe layers. Using other tools developed in-house such as Oracle Spatial based network tracing and import/export utilities, a complete hydrologic and hydraulic model can be created in the GIS on the fly.

Delineation of subcatchments and creation of hydraulic model with a DEM: the SF DPW “Urban Drainage” model
The “Urban Drainage Model” is based on the commonly used “steepest path” approach, using a gridded elevation model. The algorithm is illustrated in the figures below:


 

Study Area
Using the “Urban Drainage” model, catchment areas were delineated for the City of San Francisco (approximately 30,000 acres). A subcatchment was created for each drain in the area. Since the process is designed to be automated, revised catchments can be created whenever the source data changes. Amendments are made to the script based on feedback gathered after each revision.

A typical block section of the LiDAR DEM (the contrast has been exaggerated). The streams generated using the “Urban Drainage” approach roughly follow the line of the gutter.
 

Smoothing and Hydro Enforcement
Hydro enforcement is used to force the modeled surface flow to fit within known boundaries, or follow a known flow path. For example, the elevation model can be lowered along a river channel, to force the flow to stay in the channel. Similarly, known ridges can be represented by raising the elevation model along the ridge.

Property lines are used in the “Urban Drainage” model as high walls, forcing the flow path towards the nearest street. Street right-of-way is also used as a lower wall, so that any flow which enters the right-of-way is not allowed to re-enter a property. This simplified logic approximates the real-world conditions, where runoff from private property is piped to the sewer main. There are some cases where this logic is incorrect such as when a street slope runs counter to the sewer slope. However these cases are rare enough that they can be neglected when the catchments are used for large scale, basin-wide analysis. For detailed design of the smallest pipes, the automatically created catchments must be reviewed for these types of exceptions.

Drain inlets are also used as hydro enforcement features during the depression filling process. A depression is defined as any cell(s) where flow path reaches a dead end, rather than the edge of the map. Some depressions are artifacts of the aerial surveying process, and must be filled in. However, there are real depressions at each drain location. The ESRI grid hydrology tools treat null cells the same as the edge of the map, so a depression with a null cell at its bottom is not filled. By adding each drain to the DEM as a null cell, the surrounding depressions are left unfilled, which greatly improves the resulting catchment boundaries. The drain inlets become the loading point for the next downstream pipe segment in the model.

Some smoothing of the DEM must be done before it is used for delineation, such that the output catchments are smoother and easier to interpret. Elevation is averaged within a 3-meter circle radius using the Focal Statistics tool.

The LiDAR survey can be used to identify subcatchment parameters such as slope. SF DPW also uses multispectral imagery which can differentiate between pavement and vegetated surfaces. This data is used to estimate each catchment’s impervious area.

Hydraulic Model Creation Process from Sewer GIS
Part of the impetus for creating the catchment delineation tool was the availability of the Sewer GIS. The GIS includes every pipe and manhole in the City, and was developed with hydraulic modeling in mind. With the availability of a consistent set of small subcatchment, hydraulic models can be created quickly for any part of the City.

In the figure below, catchments created with the tool were used to create a basemap for use in a hydraulic study. All the pipes and catchments upstream of the “point of interest” were identified, and can be imported into the modeling software.

Typical basin studied using Urban Drainage model.
 

Conclusions
This has been an example of replacing older style engineering tasks with GIS automation. Automated GIS tools can streamline time consuming processes as well as help improve model quality and resolution. SF DPW staff continues to refine this model to help improve the reliability and resolution, as well as to add new features to allocate surface flows separately from individual house flows which will increase the model resolution even further. The sub-catchments delineated using these tools are also consolidated for use in lower resolution City Wide models which are used for large project planning.

BAAMA, a chapter of the Urban and Regional Information Systems Association (URISA), is a non-profit, professional organization that organizes bi-monthly educational forums, the annual California GIS Conference, and periodic technical tours on a broad range of geographic information systems (GIS) and automated mapping topics.  BAAMA is the vital organization of GIS professionals in the San Francisco Bay Region that promotes partnerships and teamwork with users of GIS technology to improve our environment and community.


Did you enjoy this topic? Check out these Channels:
Energy and Infrastructure, Government, Remote Sensing and GEOINT

Bookmark and Share


Stay Connected

Twitter RSS Facebook LinkedIn Delicious Apple Devices Android Blackberry






Recent Comments

Is This Innovation? The Latest LBS Offerings

Location sharing is back in the news as new apps try to offer better mousetraps for sharing oneís location with friends and family. We look at PlaceUs and Jink. Meanwhile two well-known players, LinkedIn and Google, add location sharing where itís not been before. Do these new and enhanced entries suggest location-based services are maturing?

Top Ten Stories from Esri’s International User Conference
Making Sense of the 2014 Esri International User Conference
Finding OGC WMS, WFS, WCS services
Survey India 2014 Focused on Infrastructure Development
Takeaways from the Esri Education GIS Conference 2014
Mercury Rising: When to Expect the “Warmest Day of the Year”
The Big Four Make Big Waves in Location Intelligence
How can I use crowdsourced data?

DirectionsMag.com

About Us | Advertise | Contact Us | Web Terms & Conditions | Privacy Policy
© 2014 Directions Media. All Rights Reserved