2016-2017 Funded Proposal
Research Area: This study is an interdisciplinary analysis of urban system resiliency to climate change that integrates methods from the social, hydrologic, and systems sciences. Speciﬁcally, this study focuses on urban ﬂooding, a common occurrence that will be exacerbated by climate change, and the beneﬁts of green infrastructure for ﬂood abatement. Instead of traditional designs that aim to merely mitigate excess runoff or its associated economic costs, this study aims to incorporate the disproportionate effect of weather events and their associated hazards on communities with less social and economic advantages, highlighting the need to integrate these communities in ﬂood abatement and resiliency planning.
There is a timely need to consider the urban system’s resilience to climate change. By the end of the next decade, 60% of the world population will live in cities; however, there are major concerns as to whether cities are protected against the projected increasing number of extreme weather events. Currently, city managers are forced to make infrastructure decisions complicated by massive amounts of data and uncertainty. That is, many climate models exist that attempt to project weather events and their vari-ations; however, none is 100% reliable, and they are at times inconsistent in their projections for a given location. At the same time, the impacts of these events are often unevenly spread among a population, i.e., socioeconomic factors such as gender, age, health, income, and access to institutional resources may play important roles in the path to recovery for dfferent citizens affected by weather events. Hence, it is critical to develop tools to distill data into a usable format, while accounting for uncertainties and maintaining a holistic, socially responsible view. The purpose of this study is to develop tools which address the complexity and uncertainty of climate projections to allow optimized choices for building resiliency into urban systems subject to various economic, physical, and social constraints. Although applicable to a wide range of urban systems, these optimization methodologies will be tested herein through the lens of urban stormwater infrastructure.
A recent global survey of 468 cities found that changing stormwater runoff patterns and stormwater management requirements are the most widely anticipated urban climate change impacts. At the same time, in its 2013 “Report Card” for American infrastructure, the nation’s stormwater systems (in combination with wastewater) were awarded a D+, indicating the poor state of these critical components of the urban landscape. The combination of these two variables is concerning for stormwater systems, where the size of pipes is selected based on historical rainfall data, rather than the changing weather patterns. Thus, climate change and the associated overwhelming of stormwater pipe systems is likely to cause increased ﬂooding in urban watersheds, escalating the already present trend of ﬂooding and ﬂash ﬂooding as (on average) the leading cause of weather-related fatalities in the U.S., beyond even hurricanes and tornadoes. In general, ﬂood events can have negative physical, emotional, and social health impacts on individuals and communities. Those most vulnerable to such impacts may include people with low income, poor health, low education levels, substandard housing, or low social or political capital.
Replacing existing stormwater sewers with pipes of larger capacity would be prohibitively expensive and time consuming in many urban environments due to surrounding infrastructure and population conﬂicts. However, building resiliency into urban stormwater systems through the use of green infrastructure (GI) is an increasing trend nationwide. The 2014 Intergovernmental Panel on Climate Change (IPCC) has identiﬁed changes to urban drainage systems as a key adaption issue for North America and recommends consideration of low-regret strategies such as GI to reduce ﬂooding while also providing co-beneﬁts to freshwater provision, ecological processes, and freshwater ﬁsh populations. As such, GI has been deemed as a way to build better infrastructure as part of the National Academy of Engineering’s Grand Challenge to restore and improve urban infrastructure. Engaging social systems and groups in GI decision-making, installation, and maintenance is a critical component of building better infrastructure and is in line with the American Academy of Social Work and Social Welfare’s Grand Challenge of creating social responses to environmental change. Although GI practices have been evaluated for their performance under climate change scenarios at the site scale, little is known about how they work in concert at the watershed scale to ameliorate ﬂooding and build resiliency to climate change, nor about how public preferences for GI may affect their adoption and sustainability.
Work Plan: We propose a new interdisciplinary approach that can signiﬁcantly improve the theoretical and practical understanding of stormwater management under climate change uncertainties. This project will be integrated with an ongoing effort by PIs Khojandi, Li, and Hathaway by providing additional data and insight into socio-environmental processes in urban watersheds. These processes are critical to under-standing the urban environment, but data are lacking to make these additions in the current efforts. To showcase the contribution of our approach, we will focus on Second Creek in Knoxville, TN, a seven square mile watershed with approximately 40% imperviousness that includes portions of UTK’s campus and down-town Knoxville. Second Creek has been established by PI Hathaway as a learning laboratory as part of the Knoxville Urban Observatory (KUO) founded by PIs Mason, Ellis, and Hathaway through prior ISSE seed funding (Figure 1). Rainfall and runoff data have been collected at Second Creek for over a year and a half, allowing a detailed understanding of system function. Further, the PIs have substantial geospatial data for the watershed including stormwater sewer locations, land use characteristics, and digital elevation models.
This project has two primary goals: (i) to better understand and model the social dynamics related to ﬂood resiliency and GI implementation in Second Creek, and (ii) to develop methods that account for social dynamics within a GI optimization tool to maximize the social, environmental, and ﬁnancial beneﬁts gained from GI implementation as a whole.
(i) Social Dynamics. Social vulnerability to the impacts of ﬂooding will be assessed through secondary data sources (U.S. Census Bureau) and through primary, community-engaged data collection in select neighborhoods within Second Creek watershed boundaries. First, a ﬂood hazard vulnerability map of the Second Creek watershed will be developed through aggregating past citizens’ complaints and/or statistical analysis on the amount of excess runoff following simulated storm events within the watershed boundaries using the Stormwater Management Model (SWMM), an open sourced software supported by the US EPA. Then, social and economic population characteristics at the census block group level will be overlaid with ﬂood hazard vulnerability, to create a more nuanced, location-speciﬁc understanding of ﬂood impact risk in Second Creek. Through a three-pronged approach to primary data—participatory research meetings with neighborhood associations, quantitative surveys, and qualitative interviews—we will also examine residents’ perceptions of ﬂood risk and impacts, and awareness and perception of GI as a ﬂood abatement strategy, all of which secondary data are currently unable to provide.
We will develop agent-based models (ABMs) to simulate the social dynamics between citizens. As a bottom-up, decentralized approach, this agent-based modeling method enables us to study a system’s global behavior as a result of interactions of individual behaviors. Through ABMs, we can gain better understanding of how citizens as a whole perceive ﬂood risk, impacts, and GI strategies; also, we can then conduct “what-if” analysis and sensitivity analysis, and obtain optimal strategies to raise the awareness of ﬂood risk and GI strategies under given constraints. The above mentioned data will be used to calibrate parameters for the agents inside ABMs.
(ii) GI Optimization Tool. The GI optimization tool will be populated with various physical, economic, and social data pertaining to Second Creek. For example, hydrologic inputs such as the portions of the watershed most prone to ﬂooding and the necessary storage to alleviate this ﬂooding will be gleaned from SWMM. Simulations will be run across a range of climate change scenarios to understand how the results will vary amongst the various climate projections, and thus what uncertainties exist in the system. Inspired by Dantzig and Beale’s classical stochastic linear program, we propose a novel stochastic model to optimize GI placement along the Second Creek. This model is able to “peek” into the future and incorporate uncertainties into the decision making process to optimize the expected costs/beneﬁts subject to various economic, physical, and social constraints. Our objective is to minimize the “overall cost” incurred due to placement of GIs and the consequences of excess runoff due to changes in precipitation level in the future (2025-2030).
Of particular interest for this project is the addition of social dynamics, whereby the effects of ﬂooding on the local population are informed by the outcomes of goal (i) and integrated into the model. Such socio-environmental considerations are critical to ensuring environmental justice and appropriate allo-cation of resources to vulnerable populations. Therefore, we interpret “overall cost” broadly with the goal of improving the stormwater management system in a socially responsible fashion. That is, the costs will be estimated as a function of land cost, damage cost due to excess runoff, and more impor-tantly, the number and socioeconomic status of the individuals affected. For instance, excess runoﬀ in one particular residential area (e.g., a neighborhood adjacent to city center) may cause less economic cost than in another (e.g., city center), but may affect individuals who have less access to formal or informal resources for coping and recovery. Hence, there is a balance that must be found whereby social and population impacts are considered in addition to more traditional or solely economic factors.