 |
| Matilija Dam, aerial view of impounded sediment, May 2024 |
Planning for the removal of Matilija Dam has been ongoing since 1998 (see chronology report). As with all dam removal projects, sediment management is the primary concern. Questions include how much sediment has accumulated, how best to remove the sediments to restore the natural stream gradient, and what are the downstream impacts when sediment is released from the former reservoir?
This last question is particularly relevant to the Ventura River where significant floodplain and water supply infrastructure has been developed since Matilija Dam was constructed in 1947. Concerns about potential impacts to aging and deficient levees downstream as well as the Robles diversion have led to a multicomponent project called the Matilija Dam Ecosystem Restoration Project (MDERP.) Although complicated by the multiple agencies, interests, and components, this complex endeavor will ultimately enhance watershed resilience through upgraded infrastructure and restored natural processes.
 |
Map of MDERP project components,
Ventura County 2024 |
Field studies have confirmed the general profile of sediments trapped upstream of the dam. Over the decades, the remnant reservoir accumulated fine silt and clay while the larger "coarse" sediment accumulated in the delta and upstream channel.
Note that over the past 3 decades since studies began, it is estimated that approximately 50% more sediment has accumulated behind the dam, primarily due to the 2018 Thomas fire and the floods of 2022-2023. Currently, sand and gravel has overlain the fine reservoir deposit and filled to the level of the outlet on the back side of the dam so that sand and gravel are now being transported through and over the dam.
Current planning for dam removal includes engineering two 12ft diameter orifices that will be opened during a flood to allow high flows to head-cut a channel upstream through the sediment (AECOM, March 2022). This initial flush will mobilize a large percentage of the fine sediments stored in the former reservoir area resulting in temporarily high turbidity (i.e. "muddy" water) downstream.
 |
| Dam removal concept for Matilija Dam |
The plans call for complete removal of trees and other vegetation upstream of the dam and excavation of a "pilot channel" to initiate head cutting through the middle of the reservoir area to maximize evacuation of the fine sediments during this initial flush. It is expected that this channel will become relatively stable and erosion will diminish once the flow cuts down to the approximate depth and alignment of the historic channel. A storm prediction analysis (AECOM, November 2022) determined the preceding conditions necessary for a 3-yr or greater flood event, equivalent to about 3,300 cfs deemed necessary to adequately establish the new upstream channel and flush the fine sediment down the Ventura River and out to the ocean.
 |
channel through sediments upstream of Matilija Dam resulting from head-cut erosion during flood
alignment based on historic streambed (Stillwater 2016) |
After this initial (phase I) flush, future floods will incrementally move more of the coarse sediment (gravel, cobble, boulders) down the Ventura river. Based on other similar dam removal projects, it is assumed that a maximum of 80% of the trapped sediment will eventually be transported downstream. In other words high terraces will exist adjacent to the restored stream channel through Matilija canyon that will gradually erode over time feeding sand gravel and cobble into the Ventura River in addition to the restored natural "background" sediment flowing down Matilija Creek.
 |
| Estimates of sediment available for transport after dam removal |
A computer model was developed to simulate the downstream response of the Ventura River to the release of sediment resulting from dam removal (Stillwater Sciences, February 2020, March 2025). The table below shows the sediment volumes applied to this computer simulation. Note that the impounded sediment represents approximately 7 times the "background" supply from Matilija Creek, which makes sense since the dam has been trapping that supply for decades. Also note that the supply from North Fork Matilija Creek (which runs up Hwy 33) is just one third of the background volume from Matilija Creek.
 |
| Sediment supply applied to computer models of the Ventura River |
The change in sediment load in the Ventura River may be conceptualized by the graph below. The current trend with dam in place is for increased sediment to pass through/over Matilija Dam until the background load from Matilija Creek is added to the current sediment transport in the river. The removal of Matilija Dam will generate a large pulse of sediment that will be transported down the river with each flood until equilibrium is reached. Note that during this time there may be 6 or 7 times the current sediment load, and in the long term equilibrium condition three times as much sediment will be transported down the Ventura River. While this will provide tremendous benefits to the riverine ecosystem, migrating endangered steelhead, and ultimately the rivermouth delta and beaches of Ventura County, some of the downstream infrastructure will need to be modified to accommodate these changes.
In order to model sediment transport, several "synthetic hydrographs" were developed based on historic flows in the Ventura River. Because the "flashy" watershed is characterized by long periods of dry conditions with occasional large winter floods, hydrographs were developed to compress time and represent those periods when sediment is moving down the river.
The "Run 2b" hydrograph represents the most rapid transport of sediment by repeating the extreme floods of 1969 back-to-back four times. The "Run 2e" hydrograph was used to help predict what may happen under a more realistic scenario. The graph below illustrates the rapid flushing of sediment from Matilija Dam and the relative magnitude of this volume compared with the background inputs from Matilija Creek and other tributaries.
The map below gives an overview of the extent of the computer model. Topography was based upon Lidar imaging to provide a basis for the channel configuration. Note that modeling is limited to a "fixed bed" so the natural scour and avulsion (channel changes) are not taken into account.
The graph below shows the maximum expected downstream changes to the riverbed and 100-year flood elevation with the removal of Matilija Dam. As one might expect, the biggest changes are in the first two miles directly downstream of the dam where the river channel is narrow and constrained. Below the Robles diversion the floodplain widens and becomes braided with a natural capacity to absorb increased sediment loads. Note that the model assumed that modifications are made to both the Robles Diversion and Santa Ana Bridge to accommodate the increases in sediment. (Santa Ana Bridge widening project was completed in 2022 and planning is underway for a high flow bypass at Robles.)
Over the course of the project, several computer modeling approaches were used beginning with a one dimensional model, the DREAM2 dam removal model, and several iterations of the 2-D model which was developed to help provide greater resolution for areas of concern. As illustrated below, all of these models show similar trends.
These results may be translated into more practical terms by projecting the predicted 100-yr water surface elevation onto the floodplain topography to visualize potential changes in flooding. Below is one example of the worst case scenario in the reach directly downstream of the dam.
The next step will be translating all of this data into a risk and liability approach that considers all downstream interests. Developing a realistic assessment of these risks will be important step toward dam removal.
Commentary:
The Matilija Dam Removal Task 2.9 2D Modeling Report summarizes more than two decades of analysis directed at determining the transport and fate of the more than 8 million cubic yards of sediment currently sequestered in the former reservoir of Matilija Dam. This unique dam removal project in a steep flashy coastal drainage is complicated by the developed floodplain downstream of the dam. The need to assess the risk of the project to infrastructure and communities downstream has required an ongoing evolution in the state of the art hydrologic and geomorphic computer simulations.
This body of work concludes that that the primary impact of dam removal will occur in the first few miles downstream of the dam within a timeframe bounded by the first two 100 year flood events (perhaps many decades or more than a century into the future). In the upper reach of the study area below the dam the floodplain is geologically constrained and the pulse of sediment released from behind the dam will temporarily increase streambed elevations and water surface elevations. This effect diminishes downstream of the Robles/Meiners Oaks area as the broad undeveloped floodplain will naturally absorb and meter out this sediment pulse through avulsion and channel migration processes beyond the capacity of this modeling effort.
The Phase II SRH-2D model improves on the prior 1D analyses. However, the Phase III SRH-2D model that was developed to add sand transport to the simulations probably overpredicts streambed aggradation. The models developed here for Matilija Dam, particularly the Phase III analysis, represent an advancement in the state of the art, therefore there is not yet a similar case where this has been applied for comparison. However, to the extent that other large dams have been removed, there is some hindsight into how such predictions have played out. One important conclusion from the referenced review of dam removal (Major et al 2017) is:
Rivers are resilient. After dam removals, rivers have largely achieved apparent stability – at least within their range of natural variability – within a few years rather than decades. Despite concerns about rapidly releasing possibly large volumes of sediment to river systems, most removals to date have shown adverse geomorphic impacts (such as complete bed-texture transformation, enhanced flood risk, or damage to fish habitat) are short-lived and have been neither as damaging nor as long-lasting as feared.
References:
DRAFT 65% Design Report for Matilija Dam Removal, AECOM, March 2022
Matilija Dam Removal 65% Design Subtask 2.9: Hydraulic and Sediment Transport Modeling in SRH-2D, Stillwater Sciences, March 2025
Major, J. J., A. E. East, J. E. O’Connor, G. E. Grant, A. C. Wilcox, C. S. Magirl, M. J. Collins, and D. D. Tullos, 2017. Gravel-Bed Rivers: Processes and Disasters, First Edition, Chapter 13
Geomorphic Responses to Dam Removal in the United States – a Two-Decade Perspective.
More information and references are available at
https://matilijadam.venturacounty.gov/
