It is the morning of the fourth day after the floods began. As I look out my window, the skies are gray and cloudy with the threat of additional rain. Cleanup began yesterday. The two days before, and the long nights in between, vanished in periods of waiting interspersed with moments of intense work and anxiety. The rain varying in intensity, the water flowing in different ways down our block, neighbors building barricades to direct the flow; a chaotic jumble of images in memory.
Now the cleanup has begun. Shoveling “chocolate soup” from a neighbor’s doorway, standing in line for rubber boots, the sound of helicopters overhead. This is all the reality of living in a flood-hit area. For someone who has spent much of their professional life working on questions of climate change, disaster risk and recovery, water, and human behavior in other regions; the direct experience of a disaster in their home location is an interesting experience. I found myself watching. My analytical brain observing, documenting, comparing what I was seeing here against the factors I’d observed in other areas and the practices that are increasingly recommended for building disaster resilience. At the same time, I was engaged in a way I’d never been before. Watching the rain and wondering just how much more was coming. Dodging small boulders as I helped a neighbor divert water shooting at his foundation. Debating whether or not we should evacuate and digging out ropes in case the flow became too strong.
We live on a hillside a few hundred meters from a small, usually dry, stream called Gregory Creek. The stream level increased quickly, not the classic flash flood with a wall of water coming down. Instead the water came in unpredictable pulses; down the channel at one point then a sudden flush coming in from a side street, later another street knee-deep with waves and whitecaps. Our view and that of everyone living on the street was very micro. We had little idea of the blockages occurring upstream and how that was redirecting the flow in unpredictable ways. Some of those blockages were new – the effects of boulders rolling down the hill and sediment deposition redirecting stream flows; some were unintended but predicable – the result of gratings at the entry of culverts becoming blocked with debris; and some were semi-deliberate – the result of upstream home owners protecting their property by building walls and levies either before or during the storm. The net result, however was that as we battled to keep water from entering houses, we had little idea of how our actions were affecting those downstream. Understanding was limited to the immediate: the sound of gravel or boulders thumping on pavement serving as a guide to the intensity of the flow, the clouds above the best indicator of rain to come. The level of water on our front steps the largest guide the risk we faced. Clusters of people coming together to address the immediate need to pull logs out of a clogged culvert or build a small levy to keep water on the street rather than in the houses. The same pattern occurring on every block in every street with little knowledge, if any, of what was happening beyond that block or street. Policemen blocking intersections, themselves bemused and unsure of much beyond the immediate area.
Amazingly the power stayed on most of the time. With phones and internet working, the micro view from our street was in strong contrast with emergency announcements across the city. The sirens along Boulder Creek would blare, updates would appear on the emergency management site, and the news began to trickle in. At the same time, we had no knowledge of what was happening two blocks up the hill or two blocks down the hill.
On the second night the power failed – for how many hours I have no idea. Working in the darkness with headlamps our view became even smaller; limited to the beam of our own lights and those worn by others on the street. Information on what was going on, which had been limited in any case, became even less as the Internet failed in waterlogged cell phones ceased to work. With our knowledge limited to what we could see within the beam of our headlamps, the wider picture became ever more abstract. As the rainfall intensity increased, people’s behaviors changed as well. Neighbors banded together in ever more tight small groups to address immediate needs. At the same time, however, the transient mostly student population caught the excitement of the night. Parties with loud music blared out of a few houses. Large batches of firecrackers were set off in sudden bursts around the neighborhood. Small groups in large cars raced through pieces of the streams sending waves against the levees people were building to protect their houses and waving to everyone in high, excited, voices. A very drunken young man stumbled by lecturing the water flow on the street and whacking at it with a stick. Cohesion and anarchy side-by-side.
Early in the morning the question of sleep became pressing. The rain was slightly lighter; the water lapping at our front step had gone down by a centimeter or so. It had done that before and then picked up again as the rain increased – so the decision to sleep felt momentous and risky. Ultimately we decided to sleep with one ear tuned to the sound of the rain in the gutters and the tone of river in our street.
SUNDAY, SEPTEMBER 15, 2013
Now I’m sitting at my computer on the Sunday afternoon, four days after the rain began. Yesterday was clear and most people focused on cleanup. Today the rains returned again. We are under a flash flood watch but the overall sense is that it’s unlikely to become as intense as it was a few days ago. It’s a time for reflection, where the pressures of day-to-day realities become slightly less and analysis can begin.
From the perspective of research on urban resilience to disaster, what do I see as evidence emerging from the events unfolding around me?
First, there are observations on human behavior. The “autonomous” responses of individuals and small groups of neighbors illustrated the critical role of community cohesion in building resilience. Where people have prior connections from school, through walking dogs together, gardening, music, or other interests, they helped each other out. This sense of community pulled many others including a large portion of the transient student community in. Many friendships were probably established by working hard with some previously unknown person late at night. At the same time however this response was not (and probably could not have been) coordinated at any level above the immediate local area. It was swarm behavior. Groups of people responding, like ants, to what they saw immediately in front of them. In many cases, it protected the individuals involved. At the same time however, the actions people took may have diverted waters in ways that increase or changed impacts downstream. The behavior did, however, overall direct water into common areas, primarily streets, and away from areas owned by individuals. As a result, the streets became rivers. Where people had fewer prior connections in the community, as with some of the large student population, empathy and the willingness to assist was far less evident and a party like atmosphere emerged.
At a higher level, beyond the sight of most of those battling to protect houses, the government response systems were working. Early warning sirens along the Boulder Creek would blare as potential threats were identified. As we later found out, some of the triggers came through the formal system (spotters monitoring rainfall levels) while others came from informal sources (a cell phone call from someone watching the floods in side canyons). For those who could access the internet, the Boulder Office of Emergency Management provided clear guidance on priorities for emergency personnel – emphasizing that they were focused on life-threatening emergencies and that all other forms of support would need to wait. To those of us on the street, the police and other emergency service presence was clear but their actual role somewhat less so. Many on the street had high expectations of immediate support – a backhoe to open a clogged drain, police regulation of people driving cars down flooded streets, etc. – many needs that would be difficult to respond to within a flood of demand. The different realities of the street and higher-level disaster response priorities coming in clear conflict: a gap of understanding that’s difficult to bridge.
Overall, several lessons emerge:
The importance of prior ties and a sense of community in enabling group response – and the often less supportive nature of responses where those are absent.
The emergent “swarm” nature of those responses – directed at immediate needs but with very little ability to recognize the larger context. This swarm behavior, however, did have the unifying driver of directing water away from private property into common property – in this case the streets. As a result, it has implications for overall flood management planning.
Beyond human behavior, the critical importance of key systems came through: The fact that electricity remained on throughout most of the flood insured that phones were working, this enabled people to keep lights on and see what they were doing, and allowed some access to the Internet and outside sources of information. This was critical for some elements of organization. It may, however, have been more important after the flood when people were able to reach out and provide assistance through wider social networks. In our case, the needs were very practical: getting someone to go down and buy rubber boots at the local hardware, checking in on the needs of friends and staff members in our organization, assuring families that we were okay. Beyond power and communications, fundamental importance of transport is clearly evident. Work to clear roads and streets near our house was essential in order for us to get the basic equipment needed to help friends cleanup. The failure of roads into the mountains and across the plains is, conversely, probably the single most important limitation slowing relief and recovery efforts for towns and individuals who live in isolated subdivisions. This disruption may well prove to be the largest and most lasting consequence of the flood. For individuals however, probably the largest losses relate to flooded homes. Here, in addition to questions over the location of houses within floodplains, some of the biggest impacts are probably due to house design. In Asia most houses are built of brick and cement. In comparison to houses constructed of wood with walls of sheet rock, they are relatively resilient to the impact of flooding. In addition, wiring and basic home equipment (water heaters, furnaces, electrical wiring and circuit boards, etc.) tend to be located high on the walls in Asian homes while they are located in the basements of most homes in the US. As a result when flooding occurs, the financial losses related to damaged equipment and buildings in the US are much higher. These losses will be a major blow for many families in our area.
Finally there is the question of conflicting institutions. A neighbor up the street framed one issue quite succinctly: “My house is next to a dry stream that has been designated as a wetland area. Since the area has been designated as wetland, we can’t cut vegetation and as a result it clogs and floods in any large storm. There is a real conflict between flood protection and wetland maintenance. What do we want?” If you want to have open “natural areas” as flood buffers, they won’t serve that purpose if, as naturally occurs, large trees grow and the floodway becomes clogged with vegetation.
In addition, at least along minor streams such as Gregory Creek, the city has done flood control work but it hasn’t come back on a regular basis and done the maintenance required to keep the area open. There are two basic conflicts here. First, it easier to find money for projects to protect areas but it’s hard to find money for regular maintenance. Government organizations tend to be structured in ways that put a premium on projects while finding it difficult to support long-term management. Management increases the long-term running costs of government but is often invisible to tax payers. Projects are highly visible with clear price tags and end dates. Second, as mentioned above, the objective of maintaining small wetland areas doesn’t always contribute to flood control in ways that protect houses or other structures built within the often very large areas where flooding can occur. To really protect against flood losses, very large areas have to be kept open. In the case of Gregory Creek, even if the city removed non-native vegetation (a standard part of it’s maintenance program) along the current channel, the growth of native species would block channels and the stream would shift in major storm events.
The second point above highlights the fact that conflicts are inherent in existing patterns of land ownership. Boulder was settled before floodplains were mapped and before statistical information on flood frequency or magnitude was available. As a result, the land is titled to private owners. While the city can regulate new construction and insurance can be used to discourage construction in the floodplain, really opening the floodplains would require removal of houses and probably for the land to become publicly owned and publicly managed. Furthermore because stream-flow patterns are braided and naturally shifting, much larger areas would need to be protected in order to really limit the impact of flooding in any given location. As a hydrologist, this was clearly evident in the patterns of sheet flow and sediment deposition that occurred throughout the flood. Throughout the heavy nights in days of rain one could hear the clatter of boulders and gravel being pushed down the street. Small embankments would grow in backwaters and wherever flows slowed down. Larger ones would form wherever trees, fences or debris blocked the flow. These in turn would divert the stream in new directions when water rose again. As a result, within the broad flow region, given current patterns of landownership effective protection of settled areas is difficult.
Overall, from the perspective of urban flood resilience, the evolving case Boulder highlights the following:
The spontaneous responses that emerge where communities are connected – but how disconnected those responses are from both the slightly larger context within a watershed and from higher-level government responses;
The fact that autonomous responses do have a common driver of diverting water away from individually owned property and into streets. This contributes to the role of streets as major flood ways;
The fragility of some key systems, particularly transport, in flood contexts contrasted with the amazing resilience of others, in this case power;
The role entrenched institutions, particularly land ownership patterns, play in limiting for flood control and wider environmental management options; and
The gaps in perception that divide responses at multiple levels – from the individuals seeking to protect their property but unaware of the consequences that these actions may have elsewhere to divides over environmental and flood protection objectives and the role of government in managing those.