By Jason Gerhardt
In 2004, a major flood in Boulder, Colorado was listed as one of six natural disasters waiting to happen in the United States (1). The week of September 9th 2013, Boulder experienced just such a flood.
While devastating in human terms, this event provides a glimpse into a world most of us are statistically unlikely to ever see again, a glimpse certainly worth using to our advantage to ponder an important question for the Front Range of Colorado and beyond: what does permaculture design have to offer in terms of flood planning? From ecological site assessment to creative design patterns, the permaculture lens provides a different vantage point, one that leads to a collection of unique design strategies.
An unsettling sense of place
Located at the base of the Rocky Mountain Foothills, the landscape is stark, with steep mountains and mesas jutting up from the Western edge of Boulder neighborhoods, and flattened landscapes stretching as far as the eye can see to the East. The mountain canyons drain directly into town and the waterways braid their way through the plains.
Most cities and towns in the arid Western US are logically sited alongside significant water sources. In the Front Range, Denver is on the banks of the South Platte River, Boulder on the banks of Boulder Creek, and Fort Collins on the Cache la Poudre. While the rivers and creeks provide much needed water during droughts, like the area has experienced for the last ten years, they are a double-edged sword. It is an odd experience to be praying for rain one day and cursing the skies the next, but the Front Range landscape is serrated and cuts deep and fast.
The mountains, foothills, and canyons exhibit characteristics of a cool-temperate climate, getting significantly more precipitation (mostly in snow) than the flatlands. The plains and piedmonts, while exhibiting some temperate qualities, are much more of an arid region with yuccas, cacti, and ephemeral weeds providing the prime indication of the need to conserve water. While the residents here are used to wet/dry fluctuations within the various microclimates, none of the usual patterns were useful the week of September 9th in 2013 when the climate resembled the tropics.
History repeats itself…sometimes
While flooding is anticipated on the Front Range, and is an obvious force by looking at the patterns of the landscape, the dry times are far more prevalent, making it quite a shock to experience tropical moisture. Boulder’s previous 100-year flood occurred on May 30th, 1894 and measured on Boulder Creek at ~13,000 cubic feet per second (cfs). 4.5 to 6 inches of rain fell on the mountains West of Boulder over a 96-hour period, during spring snow melt season, making matters significantly worse (2). This caused the most catastrophic flooding the area had ever seen, confirming the flood danger warnings of indigenous people of the area. This event still serves as the benchmark for 100-year flooding in the Boulder Creek drainage.
In 1976, a flash flood hit the Big Thompson Canyon about 20 miles North of the City of Boulder, killing 144 people and destroying hundreds of homes. Up to 14 inches of rain fell over two days from July 31st to August 1st, though the majority of the rain fell over a 4 to 6 hour period. In one location, 7.5 inches of rain fell in a little over an hour. The highest recorded stream flow rate on the Big Thompson was an astounding 31,200 cfs, nearly triple the rate of Boulder Creek’s 100-year flood benchmark in 1894 (3).
Clearly, flooding is nothing new to the Front Range landscape and human populations. The dramatic topographic relief that attracts so many to the area exists, in part, because of massive amounts of runoff and erosion. While flooding is expected in the region, each event bears differing characteristics with one regular pattern being the storms tend to target specific drainages. One of the things that makes the flood of 2013 remarkable is that it didn’t follow this previous pattern.
Clouds lifting off the rugged Boulder landscape at the end of the storm.
Engines of the atmosphere
Labeled a “100-year flood” for numerous drainages along the Front Range of Colorado, this term isn’t a very accurate way to view the significance of the storm. 100 and 500-year flood calculations are measured by taking stream flow rates in cubic feet per second (cfs), which would seem like a decent way to determine the significance of a flood. In reality, this ignores the broad landscape by focusing solely on major drainages and their floodplains. Comparing flow rates from one decade or century to the next is also making a major hydrological assumption, and an erroneous one at that, that watersheds act in the same way from flood event to flood event.
Such terms also seem to indicate the frequency of flood events, as in, such a flood will occur approximately every 100 years. In truth, the term simply means there is a 1 in 100 chance of such a flood occurring in any given year. In Boulder, Colorado’s 2013 case, flooding to this degree hadn’t been seen for nearly 120 years. It might be more accurate to call it a “historically significant rain event”.
When the rain started the evening of September 9th, there was a feeling that this storm would be different. Rain was forecasted to be in the area for days. The air was uncharacteristically heavy, the clouds extremely low, and the rain was falling perfectly vertical, all of which were out of the norm for the area. Orographic lift, or upslope flow, was streaming moisture up from the Gulf of Mexico and the Pacific Ocean, due to a low-pressure system that was stuck over the Great Basin and stubborn high pressure in the plains. It rained heavily for three to four days, stopped for 12 hours, then started again for two days. After all was said and done, from Monday to Monday, Boulder’s official rainfall was 17.15 inches, though up to 22 inches were recorded in other parts of town (4). This storm felt like afternoon thunderstorms in the tropics, but for 8 days straight, in the drylands.
What’s more was the extent of the rain. Flooding from this weather phenomenon was occurring from New Mexico to the Colorado/Wyoming border. From southern Colorado northward, Fountain Creek, the South Platte River, Coal Creek, Boulder Creek, Four Mile Creek, Four Mile Canyon Creek, Left Hand Creek, the North and South St. Vrain Rivers, Big Thompson River, Cache la Poudre River, and many others were all out of their banks and causing significant flooding. Thus began the largest air evacuation since Hurricane Katrina in New Orleans. The scope and duration of the damage was unprecedented.
Boulder County was one of the hardest hit areas, receiving more rainfall than anywhere else. Normal annual precipitation in the City of Boulder is 18 to 21 inches, which is basically what was measured during the length of this one storm. Daily, weekly, monthly, and annual precipitation records were all broken.
While the length and intensity of the rain, along with the topography of steep mountains draining into narrow creeks, was enough to make for a devastating situation, there were two other factors that contributed significantly to the flooding: wildfire and city streets.
From source to sink
Disasters of this magnitude are nothing new to this region. The only difference is the usual natural catastrophes are of the hot and dry kind. From 2010 to 2013, over 200 square miles of mountain and foothill forest, from Colorado Springs to Fort Collins, burned due to wildfire. All of this land is located above the Front Range cities and sheds water directly into the creeks and streams that run through them.
This is always the fear after the threat of wildfire has subsided—once the landscape is burned, runoff rates increase dramatically, causing flooding. One reason for this is that there are fewer trees to intercept the rain, but compounding the problem is that much of the accumulated mulch-like debris characteristic of a forest is burned away, and a hydrophobic, charred-over soil is left behind. Add the steepness of the terrain and you have the ingredients to turn a hot and dry disaster into a wet, muddy one.
While 200 square miles of burn area might seem enormous, the reality is that’s only a 3 to 4 year snapshot. The Front Range foothills are prone to fire and many small fires occur each year that aren’t captured by statistics. Past burn areas, a decade or more old, are also still largely unvegetated. It only takes a trip through one of the canyons or a climb to the top of one of the mountains to see the legacy of fire in the landscape. Unfortunately, it is a legacy that compounds the degree of disaster for years and years to come
Unvegetated burn scar at the top of the watershed, with debris catching/water diverting knee wall and terrace that saved the home to the left.
In the floodplain of your own street
The week after the flooding, I had a conversation with long-time Permaculturist Andrew Millison, out of Corvallis, Oregon. In giving him an assessment of what was happening from the flooding, I mentioned some of the worst flooding wasn’t happening along Boulder Creek, but higher up in the landscape in areas that were never much considered at risk of flood. When I told him that one area of Boulder had recorded 22 inches of rain, he quipped, “that’s when you find yourself in the floodplain of your own street”.
Often left out of the conversation when talking about landforms and the natural ecology of a place is the very ground under our feet—cities. One of the most recognizable landforms from space, we would be foolish to think the things that humans have designed aren’t also effecting natural disasters. In the case of rain, the excess of impervious surface in towns and cities can have a tremendous impact on localized flooding. Streets turn to rivers, curbs turn to stream banks, and homes suddenly become sited in a giant grid of floodplain.
And that’s exactly what happened. Residences distant from creeks found their street-sheds sending torrents against their homes and through their yards. Soils became saturated, runoff increased, and basements and crawl spaces became inundated all over the region. People outside of designated floodplains suddenly found themselves scrambling for sump pumps, sandbags, and moving valuable items to higher ground. Many also came to discover their homeowners’ insurance doesn’t cover flooding.
In the floodplain of your own street. Photo Credit: Laura Ruby
Everything was undersized
While the scope and scale of the damage is truly hard to comprehend, there seems to be one constant in it all, there was just too much rain for the size of the drainages and drains. It is this fact that needs to be focused on in order to plan for similar events in the future.
Modern water engineering would suggest making both drainages and drains bigger—to dredge the creeks and streams deeper and straighter, and be prepared to shed more water out of the towns and cities with bigger drains. That or accept the fact that these floods will occur occasionally and be prepared to count the losses.
There is another way of looking at the problem though. Instead of looking at the sink, let’s look at the source. During the flood, the City of Boulder’s stormwater drainage system (the sink) reached capacity in many places. To reiterate, there was just too much water running off the landscape (the source). The problem from looking at the source appears to be an excess of runoff rather than a deficit in drainage. As it turns out, other cities, from wet and dry regions alike, are looking at flood problems from this angle too.
Too much water is the problem AND the solution
If only there were a valve to the skies that could turn on the rain when we want it and shut it off when we don’t. Unfortunately, that’s not how the earth works, but we can achieve a similar effect by another means. Rainwater harvesting is fightin’ words in Colorado, but it may just be the panacea to our problems. From drought to flood, water harvesting works both ways. All one needs to do is look to cities like Tucson, AZ and Portland, OR.
In Tucson, city planners are looking at the excess of impervious surface in urban areas as a resource (5). From small rainstorms to deluges, the city street-sides of Tucson are becoming more and more lush. Rain gardens are being installed citywide along the streets to slow, spread, and sink the rain into the soil. This is not only helping to solve their growing outdoor water use and monsoon flooding problems, but beautifying the city, producing food, wildlife habitat, and free air-conditioning, not to mention adding more vegetation to the landscape, thereby slowing runoff before it ever hits the ground. These rain gardens are also helping prevent downstream flooding by capturing the rain as close to where it falls as possible.
Portland, Oregon is looking at rain gardens as a solution to water pollution rather than drought. Surrounded by waterways that are home to a wide variety of seafood and aquatic life, the traditional economy of the region is threatened by urban water pollution. Capturing the rain as close to where it falls and infiltrating it into the soil prevents polluted runoff from entering directly into salmon runs and other habitat. As it turns out, terrestrial organisms are much more capable of breaking down pollutants than aquatic ecosystems. The organization Salmon Safe has been encouraging the planting of these water-harvesting, water quality protecting gardens throughout the urban watershed for years now (6).
Imagine a city where every building’s multiple downspouts are passively channeling rainwater through the soil, irrigating the landscape along the way. Imagine every block lined with multiple water harvesting gardens capturing the runoff of the street itself, and using it for street tree irrigation. If you can do this, you are pretty close to seeing the solution.
Soil can hold an enormous amount of water. During the flood in Boulder, it wasn’t that all soils reached complete saturation everywhere, but more that most of the human developed landscape itself wasn’t shaped to retain water, and runoff formed more easily than it might have otherwise. If, as a matter of standard landscaping practice, the land had been shaped to infiltrate water rather than shed it, it could have prevented millions of gallons of water from rushing off into the swollen creeks and streams of Colorado.
The water harvesters adage changes just a little for flooding. If you can’t sink it, you can spread it, if you can’t safely spread it any further, at least you’ve managed to slow it. Applied citywide, the effect of this strategy becomes quite staggering and makes neighborhoods, and the region as a whole, more flood resistant.
All that water stored in the soil will also make the region more drought resistant for when the rains don’t come too. One problem during the flood was getting people to turn off their irrigation systems. A message was broadcast from the Boulder Office of Emergency Management to remind people to shut them off. I know of at least a handful of residences that didn’t have to worry about that. I’ve designed and installed dozens of landscapes that use nothing but downspouts as irrigation. The soil becomes the sponge for runoff and the reservoir for dry times. All of these gardens managed perfectly well in the flood, with no overflow, back flow, or any problems whatsoever. They were simply solutions; solutions to my clients’ water bills and to downstream flooding alike.
In essence, storm water drainage is lacking redundancy and multifunctionality. Instead of solely working on the sink of the problem, we can simultaneously work on the source. In fact, it would behoove us to work on the source before rethinking the sink, and it could be a whole lot cheaper, simpler, and multifunctional to encourage a new standard of landscaping practice rather than undertaking massive stormwater drainage reconstruction.
Water infiltrating into permeable parking lot at Naropa University.
A hydrophobic fix
There is no need to stop there either. What if we could apply the same logic to rural areas and wildlands? Dirt roads need rain gardens too, but it would be particularly beneficial to direct our attention to burn scars. The principles are the same—slow the water down, spread it out, and sink it into the soil as much as possible. The form just looks a little different.
In many burn areas, the biggest resource for this work is standing dead wood. These trees can be felled and staked out along the contour of the slope in a fish scale pattern, effectively micro-terracing the landscape. This process already happens when trees fall of their own accord, but by positioning them on contour the result can be magnified by human ingenuity and mobility. This will not only prevent excess runoff, it will also aid in the revegetation of the burn scars by storing water in the soil and allowing a place for seeds to germinate and trees to grow. Vegetation is really the end goal, but you need a way to get it established.
How Rain and Trees Interact, Illustration by: Maggie Field.
In the drainages of the burn scars, the left over charred brush can be used to construct brush weirs, built perpendicular to the direction of water flow to capture more seeds, soil, and runoff. In areas with particularly high runoff volumes, rock walls or check dams can be built in the same way as brush weirs to slow the runoff and soak it into the soil.
These solutions are nowhere near newfangled either. In Boulder County, a group called Wild Lands Restoration Volunteers is already doing some of this work (7). A drainage known as Carnage Canyon has been fully restored from fire and massive erosion already. The work just needs to be replicated, and is a longer lasting use of time and energy when compared to current burn scar treatments that Boulder County has undertaken, such as spreading straw from helicopters (8).
While these water-harvesting solutions aren’t anything we don’t already know, they are wholly underapplied. When we look through a permaculture design lens, a more wholistic perspective is achieved, allowing for the consideration of alternative solutions. If we continue to shunt precious runoff out of site (pun intended) and out of mind, we can be sure there is someone downstream that won’t appreciate being flooded out. Plus, there is always another drought on the horizon, during which we will be begging to have the rain again.
So, when the next historically significant rain event occurs, we can either scramble once again to save our homes and cities, or we can dance in the streets, from rain garden to rain garden, watching our drought and flood resistant landscapes passively functioning by the power of creative ecological design.
- Floods Predicted: http://www.livescience.com/39645-colorado-flash-floods-predicted.html
- 1894 Boulder Flood: http://www.colorado.edu/geography/extra/geogweb/bouldercreek-2/preview/
- Big Thompson Flood: http://pubs.usgs.gov/fs/2006/3095/pdf/FS06-3095_508.pdf
- Boulder Flood Rainfall: http://www2.ucar.edu/atmosnews/opinion/10250/inside-colorado-deluge
- Tucson Rain Gardens: http://watershedmg.org
- Salmon Safe Gardens: http://www.salmonsafe.org/about/our-story
- Fire Restoration: http://www.wlrv.org//index.html
- Burn Scar Treatment: http://www.dailycamera.com/boulder-county-news/ci_17792604?source=pkg
Jason Gerhardt is an ecological designer and educator based in Boulder, Colorado. He has a degree in Sustainable Design and Diplomas in Permaculture Education and Design from Permaculture Institute USA. He currently teaches permaculture design and ecology at Naropa University and operates Real Earth Design. He can be contacted at jasongerhardt (at) gmail (dot) com.