Debates—Perspectives on socio‐hydrology: Simulating hydrologic‐human interactions
2015; Wiley; Volume: 51; Issue: 6 Linguagem: Inglês
10.1002/2015wr017002
ISSN1944-7973
Autores Tópico(s)Hydrology and Drought Analysis
ResumoFlooding adversely impacts humans, their economies, and their social structures. Flooding erodes land and damages infrastructure. This motivates humans and their social institutions to take actions over time that are intended to reduce the likelihood and harm, disruption, or damage associated with such flooding events. Flooding is a common natural hazard. It is the third most damaging natural hazard globally. Our warming climate is expected to increase the likelihood, intensity, and duration of future floods through more frequent heavy precipitation, decreased catchment infiltration, and sea level rise. Our floodplain development activities are likely to increase the damages associated with such flooding. Communities are challenged to reduce these flood risks. Various combinations of structural and nonstructural flood risk reduction options are available to communities. Structural measures include improving the reliability, resistance, and resilience of existing infrastructure such as flood control dams, levees, storm water drainage systems. Dredging can increase river channel flow capacities and creating natural areas and wetlands on floodplains can make more room for storing floods. Nonstructural measures include flood forecasting and warning, evacuation planning, data exchange, enhancing institutional performance and coordination (possibly through real time simulations of flood events prior to their actual occurrence), contingency planning for disasters, insurance and legal incentives to reduce vulnerability, modifying reservoir operating rules, improving building and planning codes, designating places of temporary retreat from hazardous areas, periodic comprehensive review of flood risk reduction measures, and developing and implementing adaptive flood management policies. Can we predict which if any of these options will be implemented, and under what conditions will there be the political will to take actions that affect the risk of future flood damage? Decisions on implementing various adaptive management measures may depend in part on the willingness and ability of society to systematically monitor and respond to evolving flood risks and vulnerabilities. The likelihood of adopting any of these measures over time will largely depend on how human behavior will change in response to the potential of flooding and accompanying flood damage. Di Baldassarre et al. [2013a; 2013b; 2015] have proposed a modeling framework for looking at the coupled dynamic hydrologic-social interactions involving the management of floods on the floodplains in urban areas. Their model incorporated actions a community might take over time to reduce future adverse impacts, both economic and social, of flood events. These actions, both structural and nonstructural, were influenced by stakeholder's perceptions of flood risks, which in turn were influenced by how recent their exposure was to flooding. Their model results suggested that adaptive flood management strategies lead to more resilience communities that may be less affected by future floods than are communities that implement structural flood protection measures. The dynamic model of Di Baldassarre et al. [2015] gave them insight on possible human behavior can be inferred from the results of models that include the interactions between flooding events and human responses to those events. The challenge in building such models linking natural and social processes is to avoid biasing the model to predict the social behavior that we think should happen. Their model results reinforce what we have already learned from observations, and no doubt that is no accident. Namely, people and communities do not always do what one would expect, assuming their objectives are to reduce the risk of future flood damage. Also what is prudent in the long-run is not always what people desire in the short run—say immediately after a flood event. Their models were set up to allow these outcomes. Nevertheless their modeling framework is indeed an impressive stab at including social behavior in a hydrologic simulation model of urban flooding. The question is whether extensions of this research in what they term social—hydrologic modeling [Sivapalan et al., 2012] will eventually lead to behavioral insights we do not expect or may not have observed, and will these new insights or knowledge help us identify, create incentives for, and implement specific and improved flood management policies and practices. It would be interesting to modify and expand their model in a way that would enable estimates of the extent to which specific flood management actions could lead to social policies and decisions that reduce flood damage and flood damage reduction costs, both in the short and long run. For example, their model accounts for changes in the memory of hardship from direct flood experiences depending on the frequency of flooding. Their results suggest those experiencing more frequent flooding are more likely to prefer adaptive flood damage reduction measures than those who are less frequently impacted by floods, in part since they will be less vulnerable to flooding events in the long run. Can we therefore conclude people will implement such measures? Predicted "rational" behavior and actual behavior may not be the same. The challenge of writing mathematical modeling components that would permit estimations of how human behavior might change in response to any of these factors, or other flood risk reduction measures, is not only in inventing the variables and identifying the parameters and their relationships that would describe the possible ranges of behavior but developing the model components in such a way that does not just produce the results we should expect. In other words, human behavior can be surprising, and we would like to be forewarned about and prepare for such possible surprises. What Di Baldassarre et al. [2015] have done is to plant a seed in a garden rich in possible extensions that they and others will surely nurture and expand. The results of the model of Di Baldassarre et al. suggests that societies opting for nonstructural adaptive measures tend to be less vulnerable to future floods than are those preferring and implementing structural measures. Yet why are such actions not always taken? What influences individual human and group behavior? It is more than just how frequent flooding occurs; serving as a reminder that indeed one should be prepared to cope with floods. One can find many case studies that support the conclusion that human/social behavior is not easily predicted. I believe this is mainly because we who attempt to model these interactions and predict behavior tend not to know all the factors that lead people and their social institutions to do what they do and to make the decisions they make. But we know they make decisions that sometimes increase rather than decrease the likelihood of extreme flood events and their adverse impacts. This includes not only our combined contributions to a changing climate that increase the scale and frequency of floods due to higher-intensity rainfall and rising sea levels but also to our decisions affecting the development of flood plains that reduce their capacity to absorb flood waters. Consider for example the findings of Watson et al. [2009]. They examined the responses of individuals and communities living and working in parts of the UK that were affected by river flooding during January 2005. Some home owners chose to rebuild in flood hazardous areas rather than elsewhere. They are now equally, if not more, vulnerable to future flood risk than before the flood. Instead of making space for, and learning to live with, floods, the overall philosophy of flood management for these landowners seems to have been one of resistance and reinstating normality even though this maintained if not increased their preexisting vulnerabilities to flooding. Other communities in the UK who were affected by the extensive flooding chose different ways to reduce future flood damages. While improving and adding to existing flood banks and flood walls, they also built diversion channels that improved biodiversity and provided recreational paths for the local community. Residents of homes next to a river in one community together bought a field behind their homes and turned it into a local nature reserve and flood storage basin. The field now provides a form of flood protection and serves as a local recreational area (https://www.gov.uk/government/case-studies/). The Netherlands provides a classic example of a national strategy of making room for rivers and their floods. Despite the general consensus that more coastal and floodplain land will have to be used to store excess flood waters, it is also clear to the Dutch that major structural defenses will still be needed to protect urban areas. Relocating people, commercial, and other economic activities off urban floodplains is both politically difficult and expensive. Paying the cost of structures that provide the desired flood protection for communities located on floodplains may be most socially acceptable course of action in built-up areas. The value of the assets at risk may justify that cost. For example, more than 10 million people live in the areas at risk of extreme floods along the entire Rhine in Europe, and the potential damage from floods amounts to € 165 billion. Coastal areas of Europe are also at risk of flooding. The total value of economic assets located within 500 m of the European coastline, including beaches, agricultural land, and industrial facilities, is currently estimated at € 500–1000 billion (http://www.eurosion.org/). While all this economic development benefits humans, it also contributes to an increase in the likelihood of substantial damage during extreme flood events [de Moel et al., 2011]. In 2005, hurricane Katrina and her sister storm, hurricane Rita, resulted in many deaths and caused an estimated $150 billion in financial losses in various floodplains of the Gulf coast of Louisiana and Mississippi in the U.S. In the months after Katrina, planners in the city of New Orleans proposed to turn some of the high risk floodplain residential areas into green space that would be available, when needed, for flood storage, thereby saving money and increasing flood resilience. Community pressure by those loyal to their neighborhoods and protective of their history made that option politically untenable. Instead, the U.S. Army Corps of Engineers was given the task of protecting New Orleans from storms even more intense than Katrina. The response is a $14.5 billion "security ring" of levees and pumps around New Orleans—called the Hurricane and Storm Damage Risk Reduction System [US Army Corps of Engineers, 2010]. Hurricane Irene flooded parts of New England in August 2011, resulting in 45 deaths and $7 million in damages in that part of northeastern U.S. Just over a year later, an unprecedented weather event called Hurricane Sandy flooded parts of the New Jersey-New York coast, causing even more deaths and destruction. It was the largest Atlantic hurricane on record. The city's transportation infrastructure ceased functioning, tens of thousands were left homeless, and more were left without power for days. Many people were killed. As in the case of Katrina, the response to Irene and Sandy has been to rebuild in areas that are likely to get wet again. This reaction to rebuild on the flooded floodplains, albeit with some additional flood proofing, and the incentives, including subsidized flood insurance, that promote this disaster-by-disaster approach to rebuilding, are hard to resist, apparently. The town of Molong, Australia, has a population of just over 2500 [Australian Bureau of Statistics (ABS), 2011] and is the principal center for employment and services for the region. Floods or threats of flooding have occurred most recently in 1995, 2005, 2010, and 2012. The persistence of flooding and the ensuing damage in a relatively old and settled part of Australia suggests a failure of people and their institutions to adequately prevent damage to people's homes, assets, and livelihoods [Keys, 2006]. People in this community count on their public service and governmental institutions to provide the protection needed, even if in the past they have not been adequate. There is little adaptation to reality, little learning, and continued hope that future technical solutions and top-down actions by their governmental and social institutions will work better than has happened in the past [Alexander, 2002; Lindell et al., 2007; Phillips et al., 2010]. What do these few case studies suggest? That it remains a challenge to estimate what humans and social institutions will do in particular situations, even though it is possible, admittedly with uncertainty, to estimate the economic and social impacts of their possible decisions. Should hydrologists be trying to predict human behavior? Should they be including nonhydrologic components in their hydrologic models. I suggest if we who have some expertise in hydrologic modeling do not some other discipline will. In the recent past, some prominent hydrologists have resisted and objected to any inclusion of economic or social components linked to hydrologic processes. They were not impressed by those of us interested in creating and using economic planning models for identifying and evaluating options for water resource systems development and operation. They were critical of any models you could not calibrate and verify. It is impossible to calibrate or verify or validate a model that predicts future outcomes, and yet what occurs in the future given our actions today is what we want to know and prepare for and influence. While not denying the need for more research on natural hydrological processes, why should not hydrologists participate with engineers, planners, and economists in a collaborative effort to better understand the social as well as the economic and physical impacts of the actions we take to address future water management challenges as they unfold? Why should not we attempt, as Di Baldassarre et al. have, to incorporate the interactions of people and water in our planning and management models? I would argue that one important reason for conducting research on hydrological processes is that it might lead to improved water use, i.e., more economic and social benefits from the use and management of this resource. How to make this happen is the subject of research of many who are interested in water resources planning and management. And by necessity, part of this research must involve measures of economic and social benefits linked to hydrologic models. The goal of trying to explicitly model the interactions between hydrology, or more broadly the physical components of water resource systems and the social components or institutions is admirable. It accomplishes two important functions, in my opinion. It is again a reminder to us engineers that people matter and are indeed an essential part of our water resource systems. People are impacted by hydrological events, whether extreme or otherwise, and hydrological processes are impacted by the decisions we make in response to those impacts, namely by how we develop and manage and use our water resources and our watersheds. And we have long known that the decisions taken on how we manage this critical resource are influenced more by economic and other social factors than by changes in the statistical properties of our hydrologic records. So, this social part of our system is important. We need to understand better the dynamic, constantly changing, interactions between the social and hydrological components of our water resource systems. Given the contributions of Di Baldassarre et al., there seem now to be two modeling approaches aimed at achieving this better understanding. Their approach is to include within a hydrologic model the social components, with the hope of inferring from its results what social behaviors may happen. They have made an interesting stab at doing this, and certainly they and others will make many improvements, all with the hope I assume of predicting human behavior, eventually. Another approach is to not attempt to incorporate both hydrologic and social components in the same algebraic model, but to build an interactive interface that serves as a link between the hydrologic components being simulated for various scenarios and people—the stakeholders. The stakeholders, representing the social component of the system, can use the interface to explore what the impacts might be from taking various actions in response to certain hydrological events during the simulation and then observe how they react or the decisions they make in this simulation exercise. Admittedly what decisions are made in a simulated environment may differ from those made during and after an actual flood event, but perhaps we can still infer what behaviors are likely. This simulation approach involving stakeholders and learning from their inputs—their decisions—was one of the motivating factors in the widespread development of interactive graphs-based decision support systems in the late 1970s and early 1980s. The idea was to put people representing the social component of the system into the system and observe their behavior and their decisions. The development and application of the shared vision planning approach at the Institute for Water Resources of the Corps of Engineers is based on this idea. An application of this shared vision approach is the 5 year joint Canadian and U.S. study of the operating policies for Lake Ontario and St. Lawrence River. That study was driven in large part by social objectives and constraints. Typical of most such exercises using DSS as a means of informing stakeholders, the eventual decisions taken by the International Joint Commission (that oversaw the study and that oversees the operation of the Great Lakes) were not predictable, or anticipated, at the beginning of the study. Predicting social behavior is difficult, even after spending over 20 million dollars and 5 years working with the IJC and with stakeholders who influenced those IJC decisions [Carr et al., 2013]. Di Baldassarre et al. [2015] and the responses of Gober and Wheater [2015], Sivapalan [2015], and Troy et al. [2015] all address the challenge of incorporating social decision making and behavior in our otherwise often tidy and well tested, calibrated and verified hydrologic models. There are no laws of social behavior as there are for the physics, chemistry, and biology of water and ecology. Data on past human social behavior are no indicators of future behavior. As Gober and Wheater pointed out our views of what we want or should do in the face of some external hydrologic (or other) event is very much influenced by what we believe to be appropriate, and this in turn is influenced by what we read and hear from those we work and live with as well as from the media. The media, including the social media many of us contribute to, play a critical role in shaping human opinions and consequently our societal responses. The recent attempts by the U.S. Congress to reform and actually fix a bankrupt national flood insurance program made a lot of economic and environmental sense. But apparently not politically. The leadership of Congress, whose actions and decisions have not been modeled very successfully, feeling the political pressure from those who were adversely impacted, within 2 years passed laws that led us back to subsidizing, once again, the flood insurance premiums as well as the recovery costs of those who get wet from living on flood prone lands, often numerous times. The amounts of money received over time by some have exceeded the value of their property. This example is one of many cited by Di Baldassarre et al. [2015], Gober and Wheater [2015], Sivapalan [2015], and Troy et al. [2015], and other authors cited in their references that show the dominance of social and political processes over economic and environmental ones. While the contributions of these authors are all contributing to the development of this socio-hydrology component of our water resources research, to further this effort we will need the help of many others who are expert in other disciplines besides hydrology and engineering. Obtaining this broader perspective can only enrich and further inform our own work focused on the social as well as the economic and environmental responses to policies for developing and managing water, including the control of its extremes—droughts and floods and pollution. I have benefited from discussions with many of the authors of recent socio-hydrology papers, especially those at TU-Wien in Vienna, Austria, and the Editor of this group of papers for WRR. I thank them all for their advice and wisdom. There are no additional data associated with this commentary.
Referência(s)