Water Stewardship Information Sources

Citation BC Ministry of Forests, Lands and Natural Resource Operations. 2014. Fraser River Design Flood Level Update - Hope to Mission, FINAL REPORT. BC Ministry of Forests, Lands and Natural Resource Operations.
Organization FLNRO
URL http://www.env.gov.bc.ca/wsd/public_safety/flood/pdfs_word/Fraser_River_Design_Flood_Level_Update-Hope_to_Mission_Final_Report.pdf
Abstract/Description or Keywords The first Fraser River MIKE 11 hydraulic model was developed by UMA consultants for the gravel reach
(Laidlaw to Mission) in 2000/2001. The model was based on the LiDAR and bathymetry data collected in
1999. From 2006 to 2008, Northwest Hydraulic Consultants (NHC) was hired to develop the hydraulic
model in the sand reach (Mission to ocean) of the Fraser River. The two models were used to establish
the design profiles in the gravel reach and the sand reach respectively. Later, both models were merged
to create a single integrated model which has been used as the freshet forecasting model since 2007.
The gravel reach of the Fraser River is very dynamic compared to many rivers in North America. The
channel geometry keeps changing due to sedimentation, erosion, channel alignment shifts (avulsions)
and ongoing river engineering works such as bridges, river training structures, dredging and gravel
removal etc. These changes can result in both increases and decreases in design flood levels, so it is
necessary to re-survey and update the hydraulic model at least every ten years. In 2008, the floodplain
and the bathymetry of Fraser River were resurveyed from Hope to Mission. A completely new model
was developed in-house by FLNRO staff based on the new datasets. NHC provided guidance and
technical review of the model development work (Appendix 2).
The raw survey data included bathymetry data in the channel and LiDAR coverage of the floodplain. The
raw survey data was then processed in ArcGIS to create a surface model (TIN) to extract x-sections at the
most desired locations. The x-sections and channel network were then exported to create the new MIKE
11 one-dimensional model, which extended the full reach from Mission to Hope.
In the first phase, the model was calibrated and validated for 2007 and 2011 data respectively. The
validation showed differences near Herrling Island. It was suspected that the major channel shift
(avulsion) that took place near Herrling Island in 2009 likely triggered a significant channel geometry
change in this area. To verify this hypothesis, the reach near the avulsion was re-surveyed in 2010.
Subsequently, the surface model and the x-sections were revised in this area. The revised model showed
some improvements but the results were still not satisfactory. It was then decided to collect bathymetry
data for a longer reach to capture more of the channel changes related to the avulsion.
In 2012, the bathymetry was collected again from Big Bar to Tranmer Bar. Since the 2012 freshet was
the 5
th largest freshet in the known history of Fraser River, the bathymetry data was collected after the
freshet to capture the most recent changes that may have resulted during the freshet. In addition to the
bathymetry data, an extensive set of water level and flow data was also collected. This data included a
continuous longitudinal water level profile from Hope to Mission at near the peak flow, observed water
levels at approximately 44 gauges (including seven new continuous gauges), three sets of flow split
measurements at approximately 34 locations at varying flows and ortho-photos from Hope to the ocean
during the high flow period.
To take advantage of the data collected during a high freshet, the model was re-calibrated for the
extensive 2012 dataset. The flows used for calibration were taken from Water Survey of Canada (WSC) gauges at Hope and Harrison River. Observed WSC flows were not yet finalised at the time of calibration
and preliminary values were used. The model was validated for 2011 and 2007 data. Based on the
differences between the observed and simulated water levels at peak flows, the model accuracy is
estimated to be +0.1 m on average. Some local differences of up to 0.43 m were observed for
hydraulically complex areas such as on the left (south) bank of the river just downstream of the Agassiz
Rosedale Bridge. These differences are mainly due to the 1-D limitations of the model for areas where
the flow patterns are strongly 2 and 3 dimensional. However, overall, the model performance was
The new upper (Hope to Mission gravel reach) model was merged with the existing lower (Mission to
ocean sand reach) model to establish the new design profile. Comparison of the new profile was then
made with the old design profile. Even though the new model was created with entirely new survey data
and had a different layout than previous model, the differences between the two profiles, with a few
exceptions, were within + 0.2 m. The new profile is lower by about 0.35m near Harrison Knob (just
downstream of the confluence of the Harrison River), about 0.5m lower from Big Bar to Agassiz
Rosedale Bridge(just downstream of the bridge), about 0.2 m lower near Herrling Island, 0.2-0.3 m lower
near Wahleach and about 0.2 m higher near Peters Island.
Harrison Lake was added to the merged model to be used as the freshet forecasting model. The flow
data from a new gauge at Tenas Narrows on the Lillooet River upstream of Harrison Lake were used as
boundary conditions for testing the merged and extended model.
The new dike design profile for the Hope to Mission reach is primarily based on the new model flood
level profile. However, for short reaches where the new model profile is lower than the old model
profile and the differences are considered to be caused by changes in channel geometry, the old (higher)
water levels were adopted as the dike design level. This approach is necessary to cover the possibility of
future channel change and higher flood levels.
There are 15 dikes totalling 146 km in the Hope to Mission reach. A dike crest elevation assessment
was conducted by comparing the existing dike crest profiles with the new design profile. To ensure that
dike crest elevation data was up to date, the ministry hired a contractor in March 2014 to complete a
survey of many of the dikes in the gravel reach. This survey excluded the City of Chilliwack dikes as the
City had updated their survey in 2013.
Most of the dikes were found to have inadequate freeboard and are at high risk of over topping during a
design flood event. Within the study area, only the Matsqui B dike generally meets design level and
freeboard requirements.
Information Type report
Regional Watershed Lower Fraser
Sub-watershed if known Fraser River
Aquifer #
Project status
Contact Name
Contact Email