![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081107231508im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Final Report: Vulnerability of Low-Income Households to the Hydrologic Effects of Climate Change
EPA Grant Number: R824805Title: Vulnerability of Low-Income Households to the Hydrologic Effects of Climate Change
Investigators: Wernstedt, Kris , Austin, David , Hersh, Robert
Institution: Resources for the Future
EPA Project Officer: Winner, Darrell
Project Period: October 1, 1995 through September 1, 1998
Project Amount: $440,000
RFA: Regional Hydrologic Vulnerability to Global Climate Change (1995)
Research Category: Global Climate Change , Ecological Indicators/Assessment/Restoration
Description:
Objective:This study investigates the possible impacts on vulnerable populations?most notably lower-income households in the Willamette Basin in western Oregon?of hydrologic changes that are potentially associated with global warming. In particular, it examines three sources of uncertainties that complicate the estimation of such impacts. These are the spatial scale at which impacts are being studied, the level of social disaggregation used in a study or the degree to which subpopulations are singled out for analysis, and the characterization of vulnerability to the climate changes of interest.
The central premise of the study is that these three uncertainties can critically shape the interpretation and understanding of impacts from global warming. The primary objective of the study is to demonstrate how this may be the case. Secondarily, the study also endeavors to provide a better sense of what global warming might bring to the Willamette Basin. Finally, the study aims to illuminate several pathways by which lower-income households may be differentially affected by climate change.
To accomplish these objectives, the study focuses on three realms of possible impacts in the 30,000 km2 Willamette Basin: the physical realm, exemplified by possible changes in surface water runoff resulting from temperature and precipitation changes in the region; the economic realm, exemplified by the sensitivity of the region's agricultural sector to climate variation; and the management realm, exemplified by the potential problems that community water systems in the Willamette may face with respect to a hypothesized increase in the frequency of extreme events following from global warming. Methodologically, the study involves a mix of modeling?both simulation and statistical?and primary data collection through the administration of a survey.
Summary/Accomplishments (Outputs/Outcomes):After an introduction in Chapter 1, the project report starts substantively with a brief literature review in Chapter 2. Each of the three sources of uncertainty that are the focus of this study has been broadly addressed in the environmental change literature and, to varying degrees, the more specific climate impact assessment literature. For example, climate impact researchers have long noted the mismatch between the scale of climate drivers and atmospheric change and the scale at which impacts are experienced, yet few studies have appeared that specifically examine how the interpretation of the impacts might change as the scale perspective changes. Similarly, a broad array of studies have examined the likely unevenness of climate impacts among subgroups of the human population?most often in the context of the equity of the distribution of costs and benefits among different countries?but few have specifically examined intracountry equity or investigated the implications of changing the level of aggregation of subpopulations. In the vulnerability literature, researchers have explored a wide range of conceptual models, with Bohle, Downing, and Watts (1994) characterizing it as a function of exposure to the potential stresses of climate change, capacity to cope with perturbations, and ability to recover after a catastrophic event has passed. Viewed in this light, the intriguing question in climate impact assessments is the level of system complexity embedded in a concept of vulnerability.
In Chapter 3, we turn to our examination of water resources. Despite the relatively abundant attention paid to climate change and water resources in the Pacific Northwest and a comprehensive discussion taking place currently on the allocation of water resources in the Willamette system, there has been little, if any, specific attention to possible changes in flows in the Basin that might result from global warming. As a response to this void, we have developed a straightforward approach to examine the possible changes in flow in the Willamette under a doubled CO2 scenario. Our motivating questions are three-fold. First, is it possible to simulate flows in the Willamette with a reasonable degree of accuracy with a relatively simple model that others could use? Second, if it is possible to establish such a model, does this model perform differently when we change the scale at which it operates? And third, what do the different scale models indicate might be the effect of global warming on river discharges in the region?
Our approach for simulating Willamette flows under global warming includes a simple mass-balance hydrologic simulation model of the Willamette run on a monthly time step and based on the water-budget algorithm developed by Thornthwaite and Mather (1955). Briefly, this method estimates at a given point, or over an area, various measures of evapotranspiration, soil moisture, and soil moisture surplus (runoff) over a time series, based on time-varying temperature and precipitation inputs and location specific constants such as soil water capacity, vegetation, and latitude. After calibrating the model to base case flows, we simulate the effects of global warming by applying temperature and precipitation changes modeled by two general circulation models, the first from the Canadian Climate Centre (CCC) Model and the second from the U.K. Meteorological Office (UKMO). Consistent with the results of other studies in the Northwest on both smaller and larger watersheds, we find that average monthly wintertime flows in the Willamette are likely to increase in magnitude and be shifted earlier in winter in response to climate change under the UKMO scenario. The CCC scenario, however, generates little change in mean monthly flows except for slightly depressed flows in the first 2 months of the calendar year. When we run the same hydrologic model at a smaller scale (i.e., coarser resolution), the forward shifting of flows disappears under the UKMO scenario. Under the CCC scenario, flows generally increase over the baseline in most months of the year. Because of our monthly time step, we are not able to model directly whether peak instantaneous flows are likely to increase along with monthly flows as a result of climate change, although there is strong historical evidence that high monthly flows are strongly correlated with peak daily flows (e.g., rank correlations between peak daily flows and maximum monthly flows exceed 0.8 in all months except October at our representative gauging site in the Basin). Whether this means that low-income households may be differentially vulnerable to climate change because they are differentially vulnerable to flooding is not answerable at present. There is some limited evidence that the degree of resilience to flooding varies across counties in the Basin, and that lower-income households have less ability to cope with floods, but severe methodological constraints limit what one can say about low-income households per se.
In Chapter 4, we turn to the potential economic impacts of climate change on agriculture. We examine impacts both at a national level and, more intensively, within the Willamette. Starting with the national level assessment, we use a model from the economics literature to first demonstrate that spatial heterogeneity in precipitation and temperature is critical for estimating changes in land rents that may result from global warming. This is because of: (1) the influence of the heterogeneity in estimating the coefficients of the statistical model that relates climate and agricultural production under current climate conditions; and (2) the nonlinear responses that are observed when nonuniform climate changes (i.e., changes that vary by month and location) are imposed on regions of the country. To the extent that smaller scale analyses smooth this heterogeneity, such analyses can yield impact estimates that differ greatly from those generated with larger scale analyses that use the same underlying model.
As we move to the scale of the Willamette Basin, however, the importance of heterogeneity in climate is muted. Here, we have developed an econometric model to examine the historical relationship (1950?1993) between agricultural employment/payroll and monthly measures of soil moisture that are based on temperature, precipitation, and soil type at a very highly disaggregated level. In general, we find statistically significant evidence that previous, mid-to-late-season weather affects early season investment decisions of agriculturists. Several of these explanatory variables become statistically insignificant predictors of agricultural payroll and employment as we decrease the scale of the model, but the loss in significance is slight. When we simulate the response in payroll and employment to climate changes, we find contradictory results with our two climate scenarios: the CCC model suggests a slight decrease in payroll and increase in employment, while the UKMO model suggests a much larger increase in payroll and decrease in employment. In all likelihood, the numbers overstate the negative economic impacts and understate positive ones, since farmer behavior will adjust to the changing climate. Given this uncertainty, we can say little about whether lower-income households are more or less vulnerable to climate change with respect to agricultural impacts; however, the evidence does suggest that such households may be differentially affected by climate change. Not only do lower-income households receive a larger share of the agricultural sector's wage and sector disbursements than they do of most other sectors, but a larger share of their total wage and salary income comes from the agricultural sector than is the case for middle- and higher-income households. However, they also receive a relatively large portion of their wage and salary income from retail trade, an economic sector that would seem to be more insulated from the effects of climate change.
Finally, in Chapter 5 we examine how community drinking water systems in the Willamette are vulnerable to droughts or floods that may be associated with climate changes and what factors influence how they might respond to these extreme events. Our hypothesis is that smaller systems, our proxy for rural lower-income households, are more vulnerable to climate change by virtue of their limited resource base. To explore this, we administered a detailed open-ended survey to operators of 20 water systems that are located in the Basin. For such operators, the potential hydrologic stress associated with climate change is a minor concern, or one not recognized at all, a finding we expected. We found that water utilities managers, particularly those at smaller utilities, devoted little resources to contingency planning for extreme events under current climatic conditions; not surprisingly they see little benefit in assessing their system's vulnerability to possible climate impacts when they face other, more immediate challenges. The need to obtain and develop water rights, augmenting supplies to meet a growing demand for water, the regulatory uncertainties and costs of complying with the Endangered Species Act, and the constraints many systems face financing capital improvement projects appear in the survey responses as the major concerns of surface water utility managers. The additional burdens climate change may place on utility operators?higher wintertime flows, reduced summer stream flows, more frequent extreme events?will be felt against these current stresses.
When we turn to hypothetical changes?increases in the incidence of droughts and floods?we find a mixed message that depends on scale and definition. On the one hand, smaller utilities appear to be more vulnerable to such events. They generally have less fiscal independence from municipal budgeting processes (and thus, less ability to increase rates or to incur costs to respond to short-term disruptions), they typically lack inter-ties with other utilities that could provide buffering, and they have limited access to capital resources for intermediate or longer-term system improvements. On the other hand, many of the smaller utilities appear to have viable alternative sources of water (since they need to withdraw relatively small volumes of water) and senior water rights. Because of these senior rights, such utilities have a higher priority in exercising their permitted withdrawals from stream and river flows and therefore, greater flexibility in the case of low flow disruptions. On the flood side, we found that both large and small utilities were vulnerable to flood impacts?notably the inability to process raw water with elevated turbidity levels?although smaller systems were more susceptible than larger systems to system damage and disruption to service from extreme events. To maintain service provision and to cope with flooding, larger utilities in the past have relied primarily on stored water while smaller utilities have been able to tap into groundwater or shallow wells; however, it is unclear if this would be practical, particularly for smaller utilities, with more frequent or intense floods. Moreover, unlike the larger utilities, which typically have full-time managers and emergency plans to guide operations during flood events, smaller utilities have not devised formal procedures to plan for or to coordinate responses to flooding. Rather, they typically have sought to mobilize local resources to help them reduce the risks of flooding, a coping strategy that likely would be invisible in a small-scale analysis.
In summary, the Willamette Basin likely will not be immune to hydrologic stresses associated with climate change. The evidence from our simulation of hydrologic processes, our econometric model of behavioral responses to climate variation, and our survey of community water system operators all suggest that Willamette Basin residents and water resource managers will need to confront changes that global warming may bring. In some important ways, lower-income households are likely to be more affected by such changes, by virtue of their exposure to droughts and flooding, their more limited financial capacity to cope with the changes, and the relative importance of climate sensitive sectors such as agriculture to their income stream. However, their overall vulnerability remains very unclear, due in part to data shortcomings and methodological constraints. More broadly, knotty uncertainties concerning scale, social aggregation, and the multilayered concept of vulnerability make such a determination impossible at present. By providing extended examples of the role of these uncertainties in climate impact assessment, we hope to motivate further work in this area.
Journal Articles on this Report: 1 Displayed | Download in RIS Format
| Other project views: | All 6 publications | 1 publications in selected types | All 1 journal articles |
| Type | Citation | ||
|---|---|---|---|
|
|
Wernstedt K. Climate change: would the poor feel it worse? Resources Fall 1996;3-4. |
R824805 (Final) |
not available |
Columbia River Basin, distributional, EPA Region 10, equity, geography, hydrology, northwest, Oregon, OR, Pacific Coast, public policy, social science, socio-economic, susceptibility. , Water, Air, Geographic Area, Scientific Discipline, RFA, Water & Watershed, Social Science, climate change, Ecological Risk Assessment, Atmospheric Sciences, EPA Region, Hydrology, Watersheds, State, Global Climate Change, hydrologic models, socioeconomic indicators, climatic models, human dimension, low income households, policy making, economic models, environmental monitoring, climate models, Oregon, climate variability, land and water resources, Region 10
Progress and Final Reports:
1995 Progress Report
Original Abstract