HAZUS -MH Risk Assessment and User Group Series HAZUS-MH and DMA Pilot Project Portland, Oregon. March 2004 FEMA FEMA 436

Transcription

1 HAZUS -MH Risk Assessment and User Group Series HAZUS-MH and DMA 2000 Pilot Project Portland, Oregon March 2004 FEMA FEMA 436

2 Page intentionally left blank. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

3 ABBREVIATIONS B BIT CREW DEM Billion Building Information Tool Cascadia Region Earthquake Workgroup U.S. Geological Survey Digital Elevation Model DMA 2000 Disaster Mitigation Act of 2000 DOGAMI FEMA FIRM FIT G GIS HAZUS HAZUS-MH LCD M M MRP N/A NA NE NFIP OEM ORS PGA PMSA RLIS USGS Oregon Department of Geology and Mineral Industries Federal Emergency Management Agency Flood Insurance Rate Map Flood Information Tool Gravitational acceleration Geographic Information System Hazards U.S. Hazards U.S. Multi-Hazard Oregon Department of Land Conservation and Development Million Magnitude Mean return period Not applicable Not available Not evaluated National Flood Insurance Program Office of Emergency Management Oregon s Revised Statutes Peak ground acceleration Primary Metropolitan Statistical Area Regional Land Information System United States Geological Survey Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon i

4 Page intentionally left blank. ii Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

5 CONTENTS SECTION PAGE ABBREVIATIONS... i CONTENTS... iii FOREWORD... vii EXECUTIVE SUMMARY... ix Portland Identified Hazards of Interest... x Portland s Hazard Profiles... x Portland s Inventory of Assets... x Portland s Loss Estimates...xi Summary of Results and Conclusions...xii Uncertainties and Limitations...xiv 1 INTRODUCTION IDENTIFICATION OF HAZARDS... 3 Background Regarding the City of Portland... 3 Identification of Hazards for the Portland Study Area PROFILE OF HAZARDS...7 Earthquake... 8 Flood...10 Landslide...12 Wildland Fire INVENTORY OF ASSETS LOSS ESTIMATES Earthquake...29 Flood...37 Landslide...46 Wildland Fire...52 Portland Relative Ranking Analysis...58 Summary of Results and Conclusions...59 REFERENCES...63 APPENDICES 1 GLOSSARY 2 DATA SUMMARY 3 RELATIVE HAZARD RANKING BACKGROUND Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon iii

6 TABLES TABLE PAGE ES-1 Summary of Loss Estimates for Portland... xi ES-2 Relative Risk Categories Used for Infrastructure... xi ES-3 Summary of the Relative Risk of Loss and Exposure Estimates for Infrastructure... xii 2-1 Qualitative Hazard Ranking Results for Multnomah County Historic Landslides for Portland ( ) Inventory of General Building Stock Inventory of Critical Facilities Summary of Risk Assessment Methodology Selection Summary of the Relative Risk Categories Used for Critical Facilities Estimated Social Impacts from Earthquake Estimated Damages/Losses to General Building Stock from Earthquake Infrastructure at Risk from a 100-Year Earthquake Event Infrastructure at Risk from a 500-Year Earthquake Event Estimated Damage to Hazardous Materials Sites from PGA-specific Earthquakes Estimated Damage to Transportation Lifeline Systems from an Earthquake Estimated Damage to Utility Lifeline Systems from an Earthquake Estimated Damages for Other Lifeline Facilities from a 100-year Earthquake Estimated Damages for Other Lifeline Facilities from a 500-year Earthquake Estimated Population at Risk from Riverine Flood Estimated Exposure Values for Structures at Risk from Floods Estimated Damages/Losses to General Building Stock from Floods Infrastructure at Risk from Riverine Flood Estimated Population at Risk from Landslides Estimated Exposure Values for Structures at Risk from Landslide Hazard Infrastructure at Risk from Landslide Estimated Population at Risk from Wildland Fires Estimated Exposure Values for Structures at Risk from Wildland Fires Infrastructure at Risk from Wildland Fire Summary of Loss Estimates for Portland Summary of the Relative Risk Categories Used for Infrastructure Summary of the Relative Risk of Loss and Exposure Estimates for Infrastructure iv Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

7 FIGURE FIGURES PAGE 1-1 Risk Assessment Process Regional Map Showing the Portland Study Area and Surrounding Counties City of Portland Study Area Delineated by Census Tract Identification of Hazards of Interest Probability of Exceeding Various Peak Ground Acceleration Levels for Portland Major Past Earthquakes for the Pacific Northwest (1600 to 2002) Year and 500-Year Flood Zone Boundaries Landslide Hazard Areas for Portland Study Area and 1996 Landslide Locations Areas at Risk from Wildland Fire Hazard Distribution of General Population by Census Tract Distribution of Low Income Population by Census Tract Distribution of Elderly Population by Census Tract Annual Population Change by County for the Portland/Vancouver PMSA Change in the Share of County Population for the Portland/Vancouver PMSA Distribution of Commercial Building Stock Exposure Density Critical Facilities Transportation Lifelines Utility Lifelines Distribution of General Population Density in Relation to the Earthquake Hazard Distribution of Elderly Population Density in Relation to the Earthquake Hazard Distribution of Low-Income Population Density in Relation to the Earthquake Hazard Residential Dollar Loss Density for 500-Year Earthquake Event Commercial Dollar Loss Density for 500-Year Earthquake Event Distribution of Critical Facilities in Relation to the Earthquake Hazard Based on Soil Type Distribution of Transportation Lifeline Systems in Relation to the Earthquake Hazard Based on Soil Liquefaction Potential Distribution of Utility Lifeline Systems in Relation to the Earthquake Hazard Based on Soil Liquefaction Potential Distribution of Population Density in Relation to 500-Year Flood Extent Distribution of Elderly Population Density in Relation to 500-Year Flood Distribution of Low-Income Population Density in Relation to 500-Year Flood...39 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon v

8 5-12 Dollar Loss Density for Annualized Flood Losses for Residential Structures Dollar Loss Density for Annualized Flood Losses for Commercial Structures Critical Facilities in Relation to the 500-Year Flood Zone Transportation Lifeline Systems in Relation to the 500-Year Flood Zone Utility Lifeline Systems in Relation to the 500-Year Flood Zone Population Density in Relation to Landslide Hazard Areas Elderly Population Density in Relation to Landslide Hazard Areas Low-Income Population Density in Relation to Landslide Hazard Areas Critical Facilities in Relation to Landslide Hazard Areas Transportation Lifeline Systems in Relation to Landslide Hazard Areas Utility Lifeline Systems in Relation to Landslide Hazard Areas Population Density Exposed to Wildland Fire Elderly Population Density Exposed to Wildland Fire Low-Income Population Density Exposed to Wildland Fire Critical Facilities in Relation to Wildland Fire Hazard Areas Transportation Lifeline Systems in Relation to Wildland Fire Hazard Areas Utility Lifeline Systems in Relation to Wildland Fire Hazard Areas Projected Population Growth for Portland...62 vi Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

9 FOREWORD Hazards U.S. (HAZUS), a risk-based Geographic Information System software management tool, is a critical tool for risk assessments to support mitigation planning and other emergency management efforts. HAZUS was developed as a tool that communities across the U.S. can use to evaluate the risk and potential loss associated with priority hazards. The Federal Emergency Management Agency (FEMA) developed HAZUS Multi-Hazard (HAZUS-MH), which includes the previously released earthquake model with updated data, and new models that estimate potential losses from wind (hurricanes) and floods (riverine and coastal). In response to the requirements and deadlines of the Disaster Mitigation Act of 2000 (DMA 2000), FEMA s Mitigation Division, has conducted a series of risk assessment pilot projects across the country using HAZUS-MH. These pilot projects demonstrate the value of using HAZUS-MH to prepare the risk assessments that are needed for hazard mitigation planning. HAZUS-MH applies engineering and scientific risk calculations that have been developed by hazard and information technology experts to provide defensible damage and loss estimates; these methodologies are accepted by FEMA and provide a consistent framework for assessing risk across a variety of hazards and locations. The pilot projects apply HAZUS-MH to evaluate and analyze hazards that can be addressed or supported by the software. FEMA s Mitigation Division supports the National Flood Insurance Program and a number of other areas related to mitigation planning across a range of hazards. Mitigation focuses on taking actions to reduce the impact of future natural hazards when they occur; in turn, it protects property and human life in the event of emergencies. Therefore, mitigation is a cornerstone of emergency management. This document presents the results of the second risk assessment pilot project, which supported the City of Portland, Oregon, with short- and long-term efforts in hazard mitigation planning. The results of this risk assessment will be used by the City to address DMA 2000 requirements and meet a State of Oregon goal to develop model hazard mitigation plans in addition to the already approved Clackamas County Natural Hazard Mitigation Plan. This risk assessment document was prepared in accordance with regulations in the following sources: Oregon s Revised Statutes (ORS) Chapter 401 ( planning, preparing, and providing for the prevention, mitigation, and management of emergencies or disasters ); sample guidance outlined in the State of Oregon s 1992 Natural Hazards Mitigation Plan; FEMA s A Guide to Using HAZUS for Mitigation, FEMA 386-2, State and Local Mitigation Planning How-To Guide: Understanding Your Risks Identifying Hazards and Estimating Losses (FEMA 2001), and FEMA 433, How-To Guide for Using HAZUS-MH for Risk Assessment (FEMA 2004). The risk assessment results presented in this document will eventually be included as part of the City s hazard mitigation plan and are being applied by the City of Portland to support the identification and prioritization of appropriate mitigation strategies and actions to minimize potential losses from hazards. The State of Oregon s 2000 Natural Hazards Mitigation Plan is available at oem/programs/hazard%20mitigation/state%20plan%20volume%201.htm. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon Pilot project participants: FEMA Headquarters and Region X State of Oregon Emergency Management and the Department of Geology and Mineral Industries City of Portland Office of Emergency Management Partnering agencies, including City of Portland Bureau of Technology Services, Bureau of Environmental Services, and Bureau of Planning; Metro Regional Services; Multnomah County Office of Emergency Management; and Clackamas County Emergency Management Consultants to FEMA (Tetra Tech EM Inc. and PBS&J) vii

10 Page intentionally left blank. viii Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

11 This pilot project was conducted by the Federal Emergency Management Agency (FEMA) to assist the City of Portland, Oregon, in meeting the following needs: EXECUTIVE SUMMARY 1) Address hazard mitigation planning in response to the requirements and deadlines of the Disaster Mitigation Act of 2000 (DMA 2000) 2) Meet a State of Oregon goal to develop model Hazard Mitigation Plans The first step in hazard mitigation planning is completing a risk assessment. The risk assessment presented in this document is part of a series of pilot projects implemented by FEMA to demonstrate how the Hazards U.S. Multi-Hazard (HAZUS-MH) software program, a Geographic Information System (GIS)-based program, can be used to support the development of risk assessments as required under DMA The HAZUS-MH software program assesses risk in a quantitative manner to estimate damages and losses associated with natural hazards. HAZUS-MH can be used to streamline the risk assessment process for specific hazards because it (1) uses a consistent and defensible methodology and (2) produces maps and loss estimates that state, local governments, and the private sector can apply to develop quality risk assessments that form the basis for programs and plans required for emergency management. Hazard mitigation is any sustained action taken to reduce or eliminate the long-term risk and effects that can result from specific hazards. FEMA defines a Hazard Mitigation Plan as the documentation of a state or local government s evaluation of natural hazards and the strategy to mitigate such hazards. Risk assessment is a methodology used to assess potential exposure and estimated losses associated with priority hazards. The pilot projects demonstrate the value and benefits of using HAZUS-MH to perform the risk assessment portion of hazard mitigation plans required by DMA These pilot projects show that local communities can use HAZUS-MH as a tool to (1) make better decisions about infrastructure and disaster planning needs; (2) evaluate alternative mitigation scenarios; and (3) generate maps and other useful outputs to support community involvement and to facilitate the review of the DMA 2000 mitigation planning process and outcomes. FEMA is providing technical support to the pilot project communities, including (1) supporting local data collection and evaluation to update HAZUS-MH provided data, (2) assisting with local hazard identification and prioritization, (3) providing Executive Overview training regarding the pilot project process and HAZUS-MH, (4) training local groups to update and run HAZUS-MH using local data, (5) interpreting and analyzing risk assessment results, and (6) supporting the integration of HAZUS-MH results into a local hazard mitigation plan. This pilot project was completed in March The stakeholders for this pilot project agreed to focus on selected natural hazards and use project resources to analyze these hazards using the HAZUS-MH software and related methodology. Although DMA requirements address only natural hazards, HAZUS-MH, which is designed to assess earthquake, flood, and Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon ix

12 hurricane hazards, can also support the analysis of other types of natural and human-caused hazards. This pilot project report serves four major purposes: (1) Describes priority hazards of interest to the City of Portland and selected for this pilot project, (2) Profiles significant historic hazard events, (3) Inventories Portland s assets, and (4) Provides estimates of damages and losses or potential exposure for each hazard. Significant findings for each area listed above are summarized below. Portland Identified Hazards of Interest The study area is the geographic unit for which data is collected and analyzed. In this case, the municipal boundaries of the City of Portland comprised the study area. The initial profile ranking process generally includes a preliminary review of the hazard profiles to categorize hazards of no/low to severe concern. For Portland, the pilot project was able to build on previous ranking efforts implemented by Multnomah County. The City of Portland, as defined by the City s geographic boundary, was defined as the study area. Initially, the City identified 12 natural and human-caused hazards within the study area. From the list of 12 hazards, four were selected as hazards of interest for the City and for the pilot project. These four include earthquake, landslide, and wildland fire. The eight hazards eliminated from further consideration for this pilot project include dam failure, drought, hazardous materials site release, severe winter storm (ice and snow), terrorism, urban fires, volcano, and windstorm. Portland s Hazard Profiles The four priority hazards selected for the pilot project were profiled using historical event data, FEMA information, and state and local knowledge. Each hazard profile addresses background and local conditions, historic frequency and probability of occurrence, severity, and historic losses and impacts. Each profile includes an initial ranking of the overall importance of that hazard (ranging from low/no concern to severe concern) in the Portland study area. Of the four priority hazards, three were ranked as severe (earthquake, landslide, and wildland fire) and was ranked as high (flood). Portland s Inventory of Assets HAZUS-MH provided data and local data were used to prepare an inventory of what can be damaged or lost if a hazard occurs. Specific assets evaluated for this risk assessment include population, buildings, critical facilities, including infrastructure (roads and bridges) and lifelines (pipelines and other utilities). This report also explains how local data were used to supplement or replace HAZUS-MH supplied data. Over $59 billion of assets were identified for the Portland study area, including residential and commercial structures and building content, critical facilities, and infrastructure (utilities and transportation lifelines). x Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

13 Portland s Loss Estimates Table ES-1 summarizes the loss and exposure estimates for the Portland study area. The annualized loss column shows the average annualized loss estimated for the earthquake and flood hazard. Loss estimates could not be calculated for the landslide and wildland hazards based on current data limitations. The residential and commercial assets at risk column shows the value of building stock estimated to be at risk for each hazard based on available knowledge of the hazard and HAZUS-MH inventory data. The annualized loss estimate per $1,000 of assets at risk compares the annualized loss estimate to the assets at risk. In this case, the ratio is lower for earthquake than it is for flood. Although the earthquake hazard is estimated to have a greater loss estimate than flood ($29.7 million versus $15.4 million), its relative impact is lower because the area at risk is larger (in this case, the entire city is at risk for earthquake, whereas only flood zone areas are at risk for the flood hazard). This type of evaluation assists in comparing impacts across hazards and serves as one input to the city s mitigation planning efforts. Table ES-1. Summary of Loss Estimates for Portland Hazard Annualized Loss Estimate Residential and Commercial Assets at Risk Estimate(s) Annualized Loss Estimate per $1,000 of Assets at Risk Earthquake $29.7M $59B 0.5 Flood $15.4M $5.9B 2.6 Landslide NE $8.8B NA Wildland Fire NE $7.8B NA Notes: NE indicates not evaluated. NA indicates not available. M indicates millions and B indicates billions. Dollars rounded to the nearest hundred thousand (column 2) or million (column 3). For the flood hazard event, the 500-year at risk residential and commercial building stock is shown in column three. For earthquake, all of the residential and commercial building stock is exposed and used as the at-risk estimate. DMA 2000 requires risk to be assessed qualitatively (such as high, medium, or low) and quantitatively, where feasible. For infrastructure (critical facilities, transportation lifelines, and utility lifelines), loss or exposure estimates are presented in this pilot study as a percent of total facilities at risk or damaged. These loss and exposure percentages are then used to express relative risk qualitatively, using the scale shown in Table ES-2. Table ES-2. Summary of Risk Categories for Infrastructure Percent of Facilities Exposed or Impacted Less than 1 percent Relative Risk Classification Considered to present No or Low Risk Annualized loss is the estimated long-term value of losses to the general building stock averaged on an annual basis for a specific hazard type. Annualized loss considers all future losses for a specific hazard type resulting from possible hazard events with different magnitudes and return periods, averaged on a per year basis. Like other loss estimates, annualized loss is an estimate based on available data and models. Therefore, the actual loss in any given year can be substantially higher or lower than the estimated annualized loss. Between 1 and 5 percent Between 5 and 20 percent Between 20 and 40 percent Over 40 percent Considered to present Limited Risk Considered to present Moderate Risk Considered to present High Risk Considered to be Severe Risk Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon xi

14 The Mean Return Period (MRP) considers the severity of a hazard event that can occur within a given time period. For example, the 100-year MRP event for earthquake addresses the severity of ground shaking that has a 1 percent probability of occurring in any given year and is generally anticipated to occur at least once within a 100 year period. The 500-year MRP event is more severe than the 100- year event, but has a lower probability of occurring in any given year (0.2 percent chance of occurring in any one year and likely to occur once in a 500-year period). The 100-year MRP earthquake in Portland would be roughly equivalent to the ground shaking that would be associated with a 5.8 Magnitude earthquake occurring in the Portland Hill area or a 7.0 Magnitude earthquake occurring in the subduction zone (ocean). Table ES-3 summarizes the relative risk categories for infrastructure for each of the hazards analyzed for the Portland study area. Table ES-3. Summary of the Relative Risk of Loss and Exposure Estimates for Infrastructure Hazard Critical Facilities Transportation Lifelines Utility Lifelines Limited Risk to Limited Risk No Risk to Limited Risk Moderate Risk (2.7% to 4.3%) (0.0% to 1.4%) (3.8% to 8.6%) Earthquake 100-year MRP Event Earthquake 500-year MRP Event Flood 100-year MRP Event Flood 500-year MRP Event Landslide Wildland Fire Moderate Risk (11.6% to 16.1%) No Risk to Moderate Risk (0.0% to 7.2%) No Risk to High Risk (0.0% to 23.2%) No Risk to High Risk (0.0% to 30.0%) No Risk to Moderate Risk (0.0% to 20.0%) Limited Risk to Moderate Risk (1.3% to 18.5%) Low Risk to Moderate Risk (0.9% to 19.2%) Moderate Risk to High Risk (5.6% to 30.0%) Moderate Risk (8.7% to 13.3%) Limited Risk to Moderate Risk (1.3% to 17.3%) Limited Risk to Severe Risk (3.8% to 57.4%) No Risk to Severe Risk (0.0 to 100%) No Risk to Severe Risk (0.0 to 100%) No Risk to Severe Risk (0.0% to 45.5%) No Risk to High Risk (0.0% to 24.4%) Notes: For categories where more than one loss ratio data is included, the relative ranking reflects the highest loss ratio within that category. Green indicates no to limited risk. Yellow indicates up to a moderate risk. Red indicates high risk to severe risk. Summary of Results and Conclusions As discussed in the loss estimate section and summarized on Table ES-1, HAZUS-MH was used to support loss estimates for the flood and earthquake hazard and exposure estimates for the landslide and wildland fire hazards. For the earthquake and flood hazards, the annualized loss estimate was then compared to the total assets at risk for each of these hazards. Based on the HAZUS-MH analysis and the input of the project team, the relative ranking of the four selected priority hazards for the City of Portland is estimated as follows (from highest to lowest concern): 1. Flood 2. Earthquake 3. Landslide 4. Wildland fire The relative ranking of hazards identified through this pilot project provides a starting point for the mitigation planning committee, which will evaluate additional hazards and mitigation goals and options as it prepares the city s all hazards mitigation plan. Flood has a lower total loss estimate than earthquake ($15.4 million versus $29.7 million) but a greater proportionate impact on the assets at risk (2.6 for flood versus 0.5 for earthquake). The landslide and wildland fire hazards can not be compared directly to the flood and earthquake hazards because an annualized loss estimate was not feasible for these hazards. However, the exposures at risk for the landslide and wildland fire hazards are lower than the exposure at risk for the earthquake hazard and the likelihood that all of the inventory at risk would be lost should a landslide or wildfire occur is considered to be low. Therefore, these hazards are ranked lower than flood and earthquake for this analysis. Landslide has a greater exposure at risk xii Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

15 value than the wildland fire ($8.8 billion versus $7.8 billion); therefore, it is ranked higher than the wildland fire hazard above. Other findings include:? Growth trends indicate that the population in areas surrounding the study region is experiencing growth greater than that of the City of Portland. For example, the population of Multnomah County constituted 51 percent of the Portland/Vancouver area s total population in 1970 and only 35 percent of the total population in The City of Portland comprises a large part of Multnomah County s population. Studying growth trends can assist in state, regional and local emergency planning and mitigation decisions.? Areas along the Willamette River include flood zones, landslide potential, liquefaction potential, soft soil areas, and significant development. The multiple hazard areas along the river, combined with the level of development, appear to indicate that this area may face greater risk of losses than other areas of the study region.? Inventory in the study region is significant with, $59 billion in assets estimated, including: buildings, critical facilities, and infrastructure. Earthquake? For the 100-year MRP earthquake event, up to 1 percent of the households in the area (2,000 households) could require shelter. This increases to 4.6 percent (8,000 households) for the 500-year MRP event.? For the 100-year MRP earthquake event, minor injuries are expected to impact up to 2,500 persons and major injuries and fatalities are expected to be as high as 200. For the 500-year MRP event, this increases 240 percent (to 8,500 persons) for minor injuries and 350 percent (to 900 persons) for major injuries and fatalities, respectively.? The total economic loss ratio for residential and commercial occupancy classes for the 100- and 500-year MRP events ranged from 2 percent to 7.4 percent, respectively. The economic loss ratio represents the percent of the total building and content dollar value that would be required to repair or replace damaged structures and building content. For additional explanation of the meaning of the 100- year and 500-year MRP earthquake events, see page xii. Flood? For the 100-year MRP flood event, 4.9 percent (11,200 households) are exposed. Approximately 6.7 percent (15,300 households) are exposed to the 500-year MRP flood event.? For the 100-year MRP flood event, 29,900 persons are exposed to this hazard (that is, live within the area likely to be impacted). This increases to 38,400 persons for the 500-year MRP event. For this For additional explanation of the meaning of the 100- year and 500-year MRP flood events, see page xii or the Appendix 1 definitions for the 100-year flood and 500-year flood. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon xiii

16 hazard, socially vulnerable populations represent 17.7 percent and 22.6 percent of the total population exposed for the 100- and 500- year MRP flood events, respectively. Persons from low-income households represent 22.2 percent of the overall population. Therefore, it appears that socially vulnerable populations are not disproportionately exposed to this hazard.? Commercial building class losses account for about 37 and 43 percent of the total estimated loss for both the 100- and 500-year flood events, respectively. However, commercial buildings and content represent only about 20 percent of the total building and content value in Portland. This appears to indicate that commercial exposure may warrant further examination in relation to the flood hazard, as it appears to be disproportionately exposed to the flood hazard. Landslide and Wildland Fire At risk for the landslide and wildland fire hazards indicates the number of persons or building value lying within the geographic areas identified as defining the hazard area.? The population at risk from landslides and wildland fires is 66,400 and 64,400, respectively. This indicates that exposure is somewhat greater for the landslide hazard.? For the landslide and wildland fire hazards, total residential and commercial occupancy class value exposed to each hazard was $8.8 billion and $7.8 billion, respectively. This indicates that exposure is somewhat greater for the landslide hazard in the Portland area. Uncertainties and Limitations For this risk assessment, the loss estimates and exposure calculations rely on the best available data and methodologies. Uncertainties are inherent in any loss estimation methodology and arise in part from incomplete scientific knowledge concerning natural hazards and their effects on the built environment. Uncertainties also result from (1) approximations and simplifications that are necessary to conduct such a study, (2) incomplete or outdated data on inventory, demographic, or economic parameters, (3) the unique nature and severity of each hazard when it occurs, and (4) the amount of advance notice that the residents have to prepare for the event. These factors result in a range of uncertainty in loss estimates, possibly by a factor of two or more. As a result, potential exposure and loss estimates are approximate. These results do not predict precise results and should be used to understand relative risk. xiv Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

17 INTRODUCTION 1 In November 2002, a pilot project was initiated between the Federal Emergency Management Agency (FEMA) and the City of Portland, Oregon, to demonstrate the applicability of using Hazards U.S. Multi-Hazard (HAZUS-MH) software to address the risk assessment requirements of the Disaster Mitigation Act of 2000 (DMA 2000). The risk assessment project was conducted to evaluate priority hazards of primary concern to the community, and to estimate potential damages and losses. Figure 1-1. Risk Assessment Process This risk assessment provides a foundation for the community s decision makers to evaluate mitigation measures that can help reduce the impacts of future hazard events. STEP 1: IDENTIFY HAZARDS The risk assessment process used for this pilot project is consistent with the process and steps presented in FEMA 386-2, State and Local Mitigation Planning How-To Guide, Understanding Your Risks Identifying Hazards and Estimating Losses (FEMA 2001). Figure 1-1 shows the steps that comprise the risk assessment process. Details regarding how HAZUS-MH can be used to conduct a DMA 2000 risk assessment are provided in FEMA 433, How-To Guide for Using HAZUS-MH for Risk Assessment (FEMA 2004). Two methodologies were used to assess potential exposure and losses associated with priority hazards for this pilot project. These methodologies are summarized below. STEP 2: PROFILE HAZARD EVENTS STEP 3: INVENTORY ASSETS STEP 4: ESTIMATE LOSSES USE RISK ASSESSMENT OUTPUTS TO PREPARE A HAZARD MITIGATION PLAN? The HAZUS-MH risk assessment methodology is parametric, in that distinct hazard parameters (for example, ground motion for earthquake and discharge velocity for flood) and inventory parameters (for example, first floor elevations and building types) are modeled to determine the potential impact (damages and losses) on humans, buildings, roads and other assets. For this risk assessment, the HAZUS- MH risk assessment methodology was applied using HAZUS-MH software to estimate losses associated with the earthquake and flood hazards.? The HAZUS-MH supported risk assessment methodology was applied to analyze hazards of concern that are outside the scope of the current HAZUS-MH software. For these hazards, historic data were not adequate to support the estimation and modeling of future events and losses. Instead, HAZUS-MH inventory data and professional judgment and hazard area data regarding the geographic scope of each hazard were used to estimate exposure to the hazard. Estimated exposure to a hazard is different than an estimated loss. However, when data were not adequate to estimate loss, exposure (or at-risk inventory) was estimated as a first step to evaluating the risk. The landslide and wildland fire hazards were addressed in this manner. Over the long term, Portland will collect additional data to assist in estimating potential losses for these hazards. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 1

18 For this risk assessment, the loss estimates and exposure calculations rely on the best available data and methodologies. Uncertainties are inherent in any loss estimation methodology and arise in part from incomplete scientific knowledge concerning natural hazards and their effects on the built environment. Uncertainties also result from (1) approximations and simplifications that are necessary to conduct such a study, (2) incomplete or outdated inventory, demographic, or economic parameter data, (3) the unique nature and severity of each hazard, and (4) the amount of advance notice residents have to prepare for the event. These factors can result in a range of uncertainty in exposure loss estimates, possibly by a factor of two or more. Therefore, potential exposure and loss estimates are approximate. These results do not predict precise results and should be used to understand relative risk. This document is organized around the risk assessment process shown in Figure 1-1 and includes the following sections and appendices:? SECTION 2: IDENTIFICATION OF HAZARDS describes the study area and how the City identified priority hazards of greatest concern.? SECTION 3: PROFILE OF HAZARDS includes available data, information, and sources used to profile priority hazards.? SECTION 4: INVENTORY OF ASSETS describes the data collected and used to estimate the assets at risk (for example, buildings, people, and infrastructure, which includes).? SECTION 5: LOSS ESTIMATES describes the estimates of exposure or loss for each of the four priority hazards. It also summarizes the results of the analysis and presents annualized loss estimates, where possible, to support a relative ranking of the hazards.? REFERENCES includes a list of references.? APPENDIX 1: GLOSSARY includes a glossary of definitions.? APPENDIX 2: DATA SUMMARY provides a summary of local data collected and assessed for the pilot project.? APPENDIX 3: HAZARD RELATIVE RANKING BACKGROUND summarizes the framework for evaluating the relative risk of different hazards. 2 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

19 IDENTIFICATION OF HAZARDS 2 This section provides background information about the study area for this risk assessment. It also describes the hazard identification process. This process included identifying an initial list of hazards and then selecting hazards of interest. Four natural hazards were selected for further profiling and assessment as part of this pilot project. Background Regarding the City of Portland The study area is the geographic unit for which data is collected and analyzed. In this case, the municipal boundaries of the City of Portland comprised the study area. The City of Portland is the most populous city in Oregon and the 28 th most populous in the United States. Located in the northwest part of the state on the Willamette River, the City of Portland is the seat of Multnomah County and is in the heart of the Portland/Vancouver Oregon-Washington Primary Metropolitan Statistical Area (Portland/Vancouver PMSA). The Portland/Vancouver PMSA has a land area of 4,371 square miles and comprises six counties: Multnomah, Clark, Clackamas, Columbia, Washington, and Yamhill (shown on Figure 2-1). Its population of 1.95 million is expected to grow to 2.2 million by 2010 (Portland Development Commission 2003a). The land area for all of Multnomah County, in which most of the study area lies, is approximately 465 square miles or 297,600 acres. Multnomah is the most urban of the counties in the Portland/ Vancouver PMSA. The land area of the Portland study area is square miles or 92,672 acres. Figure 2-1. Regional Map Showing the Portland Study Area and Surrounding Counties N Grays Harbor Mason King Clark Chelan N Kittitas Thurston Pierce Portland Study Area Pacific Lewis Washington Wahkiakum Cowlitz Clatsop Columbia Clark Clackamas Skamania Multnomah Yakima miles Klickitat Tillamook Washington Hood River Multnomah miles Lincoln Yamhill Polk Marion Clackamas Wasco Sherman Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 3

20 As agreed upon by the pilot project participating agencies and shown on Figure 2-2, the study area for this risk assessment is limited to the geographic boundaries of the City of Portland. For this pilot project, the Portland study area included all census blocks and tracts entirely within the City boundaries and all census blocks and tracts that intersect with the City s boundaries. Figure 2-2. City of Portland Study Area Delineated by Census Tract N miles A major disaster declaration is a postdisaster status requested by a state s governor when local and state resources are not sufficient to meet disaster needs. It is based on a damage assessment, and an agreement to commit state funds and resources to the long-term recovery. The event must cause impacts that are clearly more than the state or local government can handle alone. Portland is subject to substantial natural hazard risks. An examination of national data on disaster declarations provides a reference point for understanding the potential for future hazard events and the potential impact that hazards can have in this area. Of the 1,037 major disaster declarations in the 50 states, the District of Columbia, and nine U.S. territories between 1972 and 2000, the State of Oregon has claimed 12, ranking it 22 nd in the number of disaster declarations for any state or territory (FEMA 2003a). Total aggregated losses from natural disasters in Oregon have mounted into the hundreds of millions of dollars during the past decade. In fact, over $220 million was provided to Oregon under several federal relief programs for three flood and landslide disasters that occurred in 1996 and 1997 (Oregon Office of Emergency Management [OEM] 2000). The Portland study area, because of its setting, is subject to natural hazards, which result in disasters when they destroy human development or cause injury. During the winters of 1996 and 1997, the Portland area experienced floods, landslides, ice storms, and other disasters. Seismic activity, heavy precipitation, weather extremes, and geography will continue to result in earthquakes, floods, and landslides. In addition, periods of long dry summers and fuel accumulation (tree, grass, and understory growth) can contribute to the potential for wildland fires. 4 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

21 Identification of Hazards for the Portland Study Area The City of Portland initiated hazard identification efforts as part of its mitigation planning process. The City of Portland identified a preliminary list of hazards of concern during a pilot project meeting in November The preliminary list was based on past planning efforts and incorporated the professional knowledge of local, state, and federal planning and emergency response personnel. During the pilot project, the project participants, including representatives of the City of Portland, county and regional representatives, the State of Oregon, FEMA, and FEMA s consultants, discussed the initial list of hazards. Four hazards were selected for further evaluation as part of this pilot project. These hazards were considered hazards of interest of primary concern for the Portland study area. Figure 2-3 summarizes the initial hazard identification and selection process conducted for this pilot project. This figure is modified from Worksheet No. 1 of FEMA 386-2, State and Local Mitigation Planning How-To Guide, Understanding Your Risks Identifying Hazards and Estimating Losses (FEMA 2001), which provides additional information on the hazard identification and selection process. Figure 2-3 summarizes the 12 initial hazards identified and the four hazards of interest selected for further analysis. It also shows historical event data and sources identified and used for the pilot project. Hazards of interest are those hazards that are considered most likely to impact a community. These are identified using available data and local knowledge. Figure 2-3. Identification of Hazards of Interest This figure summarizes the pilot project hazard identification and selection process. Column A indicates the initial hazards identified. Column B shows hazards of interest carried forward for further study based on group discussion. The table also summarizes available information regarding each hazard and its impact on the Portland study area. A B Dam Failure Drought Earthquake Flood Hazardous Material Release Landslide Severe Winter Storm Terrorism Urban Fire Volcano Wildland Fire Windstorm Summary of Hazards of Interest for Portland Pilot Project Hazard Earthquake Flood Year(s) 1872 to to 1999 No. of Events major Landslide Wildland Fire 1900 to Significant Impacts 17 deaths during past 100 years $2 billion losses (Nisqually 2001) 1948 (31.6 feet above flood stage level) 1996 (30.2 feet) 5 deaths, 100s of homes, over $1 billion loss 17 homes destroyed, 64 badly damaged, no serious injuries Area wildfires have occurred in national forests south and west of Portland but no events have caused Portland losses Available Data Sources and Maps Cascadia Region Earthquake Workgroup (CREW) 2003; Metro 1999 Metro 1999 Oregon OEM 2000 USFS 2004 Notes: Modified from FEMA 386-2, Worksheet No. 1 (FEMA 2001). Major flood events for the Columbia and Willamette River basins are indicated. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 5

22 The Portland study area comprises a large portion of Multnomah County. The County has conducted previous mitigation planning efforts, including a qualitative ranking of hazards. Table 2-1 shows an initial qualitative ranking analysis for the pilot project hazards of interest that was as part of the Multnomah County Hazard Analysis effort. This analysis was developed under a separate study but is presented here because it supported the selection of hazards for the pilot project and presented a strong starting point for further study of the area s hazards and refinement of the rankings to address the specific needs of the City of Portland. Also, this qualitative method of ranking is one option discussed in FEMA and illustrates how one county applied the state s qualitative ranking approach to initial hazard ranking. Section 3.0 provides detailed profiles of the hazards listed below. Also, the pilot project team identified additional factors to support analysis and comparison of the hazards and assigned a risk gauge initial profile ranking to each hazard, as discussed in Section 3.0. Table 2-1. Qualitative Hazard Ranking Results for Multnomah County (Multnomah County 2002) Hazard Severity Score History (2) Vulnerability (5) Maximum Threat (10) Probability (7) Total Score Earthquake HIGH Flood MEDIUM Landslide HIGH Wildland Fire HIGH Notes: Categories were assigned weighting factors as noted in parentheses under the column heading. The categories are defined as follows: Numeric values (ratings used for History, Vulnerability, Maximum Threat, and Probability): Low = 1 to 3 points; Medium = 4 to 7 points; High = 8 to 10 points. History indicates the record of previous major occurrences. (Low = 0 to 1 event per 100 years; Medium = 2 to 3 events per 100 years; High = >4 events per 100 years) X Weighting Factor of 2. Vulnerability indicates the percentage of population and property at risk. (Low = <1% affected; Medium = 1 to 10% affected; High = >10% affected) X Weighting Factor of 5. Maximum threat indicates the maximum percentage of population and property at risk under a worst-case scenario. (Low = <5% affected; Medium = 5 to 25% affected; High = >25% affected) X Weighting Factor of 10. Probability indicates the likelihood of occurrence within a given period of time. (Low = > 1 chance per 100 years; Medium = > 1 chance per 50 years; High = > 1 chance per 10 years) X Weighting Factor of 7. Total score ranking is a quantitative measure relates to the qualitative description or severity score based on the following scale: 24 to 72 = Low; 73 to 168 = Medium; and 169 to 240 = High. 6 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

23 PROFILE OF HAZARDS 3 This section presents hazard profiles for the four hazards identified in Section 2.0: earthquake, flood, landslide, and wildland fire. For the Portland study area, the City of Portland, its stakeholders, and FEMA consultants, conducted considerable research to obtain profile information for the priority hazards. Much of the data for these profiles was obtained from various regional studies including reports from the Oregon Department of Geology and Mineral Industries (DOGAMI), the Natural Hazards Mitigation Plan prepared by Oregon OEM, regional planning documents from Metro, emergency operations information from Multnomah County, and studies from the City of Portland. The Multnomah County qualitative ranking was used as one input to develop hazard profiles for this pilot project and this information was supplemented during the pilot project. Each hazard profile includes available hazard data regarding: (1) background and local conditions; (2) historic frequency and probability of occurrence; (3) severity; (4) historic losses and impacts; and (5) designated hazard areas. Available data sources are described and hazard area maps are provided with each profile. A summary of risk factors for each hazard profile includes (1) a column summarizing the Multnomah County Hazard Analysis qualitative results from Table 2-1 and (2) a column listing other summary considerations. Finally, for each hazard, the project team prepared a preliminary overall assessment of the relative risk of each hazard in the Portland study area (ranging from no or low concern to severe concern). This assessment is presented as a hazard risk gauge in the upper corner of each hazard profile. The hazard risk gauge considers all of the information obtained and risk factors considered for these profiles. It builds on, and refines, the county-wide findings presented in Table 2-1 and presents the initial ranking assigned by the pilot project team using additional information and input and focusing on the specific study area defined by the city boundaries. Hazard profiles focus on answering the question: How bad can it get? (FEMA 2001). To answer this question, data regarding historic events, future probability, severity, and hazard location are collected and evaluated. Risk factors are characteristics of a hazard that contribute to the potential losses that may occur in the study area. The hazard risk gauge is a graphic icon used during the initial hazard ranking process to convey the relative risk of a given hazard in the study area. The scale ranges from green, indicating relatively low or no risk, to red, indicating severe risk. The risk assessment outputs developed as part of this pilot project (see Section 5.0) will inform the mitigation process and may lead to new conclusions about risk and to redistribution of mitigation resources. Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 7

24 Earthquake Multnomah County Hazard Analysis Summary of Risk Factors Severity Score High Period of occurrence: At any time History (2) 20 Probability of event(s): Highly likely Vulnerability (5) 50 Warning time: No warning time Maximum Threat (10) 100 Major contributor(s): Highly active seismic zone, local soil characteristics Probability (7) 70 Cause injuries? Yes, and risk of death Total Score 240 Potential facilities shutdown? 30 days or more Magnitude (M) is a measure of earthquake size that represents the amount of energy released by an earthquake. Energy release increases 30 times for each integer on the magnitude scale. Moment Magnitude is a direct measure of energy and is a more accurate measure of the true strength or intensity of an earthquake. EARTHQUAKE HAZARD PROFILE Background and Local Conditions There are several different sources for hazardous earthquakes in the Pacific Northwest. Oregon sits on the Cascadia Subduction Zone where the Pacific / Juan de Fuca Plate is sliding under (or being pushed under) the less dense North American Plate. While earthquakes along this zone occur infrequently (none since records have been kept), plate movement can produce major earthquakes. In addition, the western part of Oregon is underlain by a large and complex system of faults (for example, the Portland Hills) that can produce significant and more frequent earthquakes. Historic Frequency and Probability of Occurrence The Metro 1999 study cites research indicating that major geologic structures capable of magnitude (M) 7 earthquakes underlie the Portland study area. Since 1820, 7,000 earthquakes have been documented in Oregon. Fifty-six significant earthquakes occurred in or near the Portland study area between 1872 and Severe local earthquakes occurred in 1877, 1880, 1953, 1962, and 1993 (Metro 1999). Strong Pacific Northwest earthquakes also include an 1872 M 7.4 North Cascades event, an M 6.8 earthquake in 1873, a 1949 M 7.1 event near Olympia, Washington, a 1965 M 6.5 event in Seattle-Tacoma, and a 2001 Olympia, Washington event that caused over $2 billion in property damage (Oregon OEM 2000). Regional earthquakes, such as the deep, intra-plate Nisqually Earthquake of 2001 (Olympia, Washington) are felt widely in northwest Oregon. Figure 3-1 illustrates the annual probability of exceeding a range of peak ground acceleration (PGA) levels in Portland. Severity There is a direct relationship between a fault s length and location and its ability to generate damaging ground motion. In Portland, smaller, local faults produce lower magnitude quakes, but their ground shaking can be strong and damage can be high as a result of the fault s proximity. In contrast, offshore or distant subduction zone quakes can generate great magnitudes, but because of their distance and depth, may result in only moderate shaking in the Portland study area (Metro 1999). The Cascadia Subduction Zone fault could produce an earthquake of M 8.0 to 9.0 or greater. Geologic evidence shows that earthquakes of similar magnitude have occurred on average every 500 to 600 years in this area. Based on the Mutlnomah County analysis and pilot project data gathering and review, this hazard was given an initial profile ranking of severe. Historic Losses and Impacts Damage results from earthquakes because structures that cannot withstand the shaking, are situated on ground that amplifies shaking, or are located on soil that is subject to liquefaction. Structures can cause injury or fatalities and suffer content and functionality losses. The 2001 Nisqually event caused over $2 billion in losses. The two 1993 Klamath Falls earthquakes (M 5.9 and 6.0) caused damage to more than 1,000 buildings and $10 million in losses (DOGAMI 2002). Since 1872, there have been about 25 damaging earthquakes in Washington and Oregon (CREW 2003). 8 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

25 EARTHQUAKE HAZARD PROFILE (continued) Designated Hazard Areas The entire Pacific Northwest is subject to the earthquake hazard. However, certain local conditions can mitigate or amplify the effects. Figure 3-2 illustrates that the Portland study area has experienced earthquakes with various intensities of ground shaking. The figure shows major past earthquakes by moment magnitude. Figure 3-1. Probability of Exceeding Various Peak Ground Acceleration Levels for Portland Peak Ground Acceleration [g's] (Soil B) Minimum PGA Maximum PGA Average PGA PGA [g] Shaking Level Expected Damage Negligible None Very Light Very Minor Light Minor Moderate Minor - Moderate Moderate - Strong Moderate Strong Moderate - Extensive Violent Extensive Soil B is characterized as a stiff, rocky soil. Portland generally has softer soil, which tends to amplify ground motion and increase potential damage Annual Probability of Occurrence Note: PGA indicates peak ground acceleration. There is a 2 in 1,000 probability of an earthquake with a PGA of 0.2 or greater in any one year (or a 2 percent chance over a 10 year period). Figure 3-2. Major Past Earthquakes for the Pacific Northwest (1600 to 2002) King N Grays Harbor Mason Chelan Kittitas Pierce Thurston Major Past Earthquakes by Moment Magnitude miles Pacific Tillamook Lincoln Wahkiakum Clatsop Polk Washington Yamhill Columbia Marion Note: The Portland study area is highlighted in blue. Cowlitz Clark Lewis Multnomah Clackamas Skamania Hood River Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 9 Wasco Yakima Klickitat Sherman

26 Flood Multnomah County Hazard Analysis Summary of Risk Factors Severity Score Medium Period of occurrence: October through April History (2) 20 Probability of event(s): Highly likely Vulnerability (5) 25 Warning time: Maximum Threat (10) 60 Major contributor(s): 0 to 3 hours (tributaries) or days (main stem of rivers) Intense precipitation, increase in impervious surface, vegetation loss Probability (7) 42 Cause injuries? Yes, and risk of death Total Score 147 Potential facilities shutdown? 30 days or more FLOOD HAZARD PROFILE Background and Local Conditions Flooding results when rain or snowmelt creates water flows that exceed the carrying capacity of river channels or other watercourses and storage facilities (for example, reservoirs). Flooding poses a threat to safety and can cause severe damage to public and private property. In Oregon, flooding is most common when storms from the Pacific Ocean bring intense rainfall (typically between October and April). The area s major rivers include the Columbia, Willamette, Clackamas, and Tualatin. There are also many streams in the area that drain to these rivers and can exacerbate riverine flooding primarily during prolonged wet periods. Local drainage flooding occurs on smaller streams, creeks, and drainage ways, and is more likely to result from heavy local storms and debris-clogged storm drain systems. This pilot project focuses on riverine flooding, which generally impacts larger areas at greater depths than drainage flooding. Historic Frequency and Probability of Occurrence Floods are a common and widespread hazard in Oregon. The Portland/Vancouver PMSA has been subject to major floods throughout recorded history. Flooding can be aggravated when heavy rains are accompanied by snowmelt and frozen ground. It was the combination of these factors that caused recent, disastrous floods in February and November Flood risk or probability can be expressed by frequency of occurrence. It is measured as the average recurrence interval for a flood of a given magnitude and can be stated as the percent chance that a flood of a certain magnitude or greater will occur in any given year. FEMA s National Flood Insurance Program (NFIP) is based on the risk associated with a 100-year or base flood; that is, a flood that has a one percent chance of occurring in any year. Although flash flooding is not a particular concern in Portland, rapid onset flooding can occur along tributaries to major rivers (for example, Johnson Creek). Major rivers such as the Willamette and Columbia will generally have more warning time before a flood occurs; these rivers are controlled by dams and also are monitored closely during heavy rain events. Severity Flooding can be a frequent, costly, and deadly hazard facing Portland; flash flooding also poses a significant danger. Many roads run through low-lying areas that are prone to sudden and frequent flooding during storms. Motorists often attempt to drive through barricaded or flooded roadways. Because it takes only 18 to 24 inches of water moving across a roadway to carry away most vehicles, this presents a significant potential cost and human health risk. The second largest impact on injuries results from people walking or playing in or near flooded areas. The lowest risk to humans is associated with flooding in the home. Generally, floods kill people in two ways: when people ignore basic safety precautions (such as evacuations and warnings), or when a flash flood hits an area with no warning. Floods can be very damaging, and depend on the depth and velocity of the floodwaters. During a severe event, buildings can be washed off their foundations; however, most flood damage is caused by water saturating materials that are susceptible to loss (for example, wood, insulation, and furnishings). 10 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

27 FLOOD HAZARD PROFILE (continued) Historic Losses and Impacts Significant historic flooding has been recorded for the Willamette and Columbia River basins in 1861, 1880, 1881, 1894, 1909, 1913, 1927, 1928, 1942, 1946, 1948, 1961, 1964/1965, and 1996 (Oregon OEM 2000). Historic flood inundation levels for the Willamette River at Portland occurred in 1894 (35.1 feet above flood stage warning), 1948 (31.6 feet), 1964 (29.8 feet), 1974 (25.7 feet), and 1996 (30.2 feet) (Metro 1999). Major past events include floods in 1948 on the Lower Columbia River in the Portland/Vancouver PMSA that caused about 25 deaths; in December 1964 and January 1965 that forced the evacuation of thousands, destroyed scores of bridges and secondary roads, caused the Willamette River at downtown Portland to have a flood stage of 29.8 feet, caused $157 million in damages, and caused 17 deaths; and statewide floods in February 1996 that caused five deaths, forced thousands into shelter, destroyed hundreds of homes, caused damages in excess of $280 million, and forced the City of Portland to erect makeshift barriers to prevent flood waters from moving into the downtown area (Oregon State Police 2003). Twenty-seven counties, including Multnomah, were eventually covered by a disaster declaration due to the 1996 floods (Oregon OEM 2000). Many residents who have suffered damage rebuild in the same vulnerable areas, only to be flooded again. These properties are termed repetitive loss properties, and are troublesome because they continue to expose lives and property to flooding (Clackamas County 2002). Designated Hazard Areas According to NFIP, Oregon has 256 flood-prone communities, including all 36 counties. Flood hazard areas are defined as areas that would be inundated by a flood of a given magnitude. The areas subject to riverine flooding have been mapped by FEMA under the NFIP and are illustrated on Figure 3-3 for the Portland study area for the 100-year and 500-year flood zones. These areas are determined using statistical analyses of flood discharge data and hydraulic and topographic analyses. A 100-year flood has a 1 percent chance of being equaled or exceeded in any one year. This flood event is also referred to as the base flood. A flood that has a 0.2 percent chance of being equaled or exceeded in any one year is called a 500-year flood. The Columbia River and Willamette River are shown below. The 100-Year Flood Zone represents the area that has a 1 percent probability of being flooded in any given year and a 100 percent probability of being flooded within 100 years. The 500- Year Flood Zone includes a greater area, but has a lower probability of being flooded in any given year (0.2 percent annual probability of being equaled or exceeded). Figure Year and 500-Year Flood Zone Boundaries N Columbia River Willamette River 100-year 500-year Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 11 miles

28 Landslide Multnomah County Hazard Analysis Summary of Risk Factors Severity Score High Period of occurrence: Fall, Winter, and Spring History (2) 20 Probability of event(s): Occasional Vulnerability (5) 50 Warning time: Hours to days Maximum Threat (10) 100 Major contributor(s): Topographic characteristics, terrain, development and construction practices, and water saturation Probability (7) 70 Cause injuries? Yes, and risk of death Total Score 240 Potential facilities 30 days or more shutdown? LANDSLIDE HAZARD PROFILE Background and Local Conditions Landslides are part of the natural, on-going process of smoothing topographical high points. Landslides occur when gravitational forces associated with slide mass exceed the resistance produced by the material holding that mass in place. Landslides are downhill or lateral movements of soil and rock that can include rock falls, slides, slumps, lateral spreading, earth and mudflows, and settlement. Landslides can result from ground saturation after intense or prolonged rainfall, erosion associated with surface water runoff, improper or poorly designed drainage systems or slopes, vegetation removal by land clearing, and shocks or vibrations from earthquakes. After wildland fires, land is also more subject to landslide because resistance forces produced by roots associated with trees, shrubs, and grass are reduced. Many hillsides in the Portland/Vancouver PMSA are unstable and vulnerable to landslides and mudflows. Landslides associated with rainfall tend to be relatively smaller; earthquakeinduced landslides may be much larger. The pilot project focuses on rain-induced landslides. Historic Frequency and Probability of Occurrence The Portland study area has been subject to major and minor landslides. Hundreds of landslides (as many as 800) were recorded during the February and December 1996 flood events. In general, landslide recurrence intervals are highly variable. Some large landslides are continuous and slow moving. Others are triggered by acute conditions and occur sporadically. Table 3-1 lists the types and number of landslides in the City of Portland from 1996 to Severity Existing mitigation and emergency directives for this risk in the Portland area evidence the high risk of this hazard. For example, the State of Oregon has a Debris Avalanche Action Plan that directs state agencies to seek solutions to reduce the loss of life from debris flow and landslides. The Multnomah County Hazard Analysis considered this hazard a high risk. Similarly, this hazard is considered a severe risk based on the Multnomah County hazard analysis and data reviewed by the risk assessment team as part of this project. Historic Losses and Impacts Hundreds of landslides occurred during the February and December 1996 flood events and accounted for 20 percent ($13 million) of the $64 million in damages associated with the February 1996 storms. During those events, 17 homes were completely destroyed and 64 were badly damaged due to landslides (Oregon OEM 2000). During the 1996 landslides, eight deaths were recorded statewide. During a March 1972 landslide, three motorists were injured in a mud and rockslide on Interstate 5 near Portland. Losses for the State of Oregon generally average less than one or two lives per year and between $1 million and $10 million annually (Oregon Department of Land Conservation and Development [LCD] 2003). Designated Hazard Areas Although the total area of land subject to a high potential for landslides is small, the consequences are serious when structures, roads, or lifeline systems are affected. Many hillsides in the study area are unstable and subject to slides and flows. Landslide losses most likely will increase because city -wide development is occurring on and near increasingly less 12 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

29 LANDSLIDE HAZARD PROFILE (continued) stable land. According to a study of the February 1996 storm, changes to slopes through cutting or filling increased the risk of landslides in 76 percent for the inventoried landslide areas in the Metro region (Burns and others 1998). The study also found that there are four dominant landslide areas: the West Hills Silt Soil Province; the debris flows in the Valley Bottoms Province along the Columbia River; the steep bluffs along Rivers Province on the Willamette and Clackamas Rivers, and the fine-grained Troutdale Formation Province (which was not analyzed for this pilot study). It is important to note that hazard maps only provide a general indication of landslide hazards. Figure 3-4 shows the dominant landslide hazard areas in the Portland study area as well as the locations of the 1996 landslides. Table 3-1. Historic Landslides for Portland ( ) (DOGAMI 2002b) Historic Landslide Type Number of Occurrences In the City of Portland Debris Flow 13 Debris Slide 56 Earth Flow 168 Earth Flow / Debris Flow 3 Earth Flow / Rockfall 1 Rockfall 11 Rockfall / Earth Flow 2 Rockfall / Mudflow 1 Slump 49 Slump Earth Flow 89 Slump Earth Flow / Debris 7 Slump Earth Flow / Rockfall 1 Slump / Debris Flow 1 Translational (horizontal movement) 1 Figure 3-4. Landslide Hazard Areas for Portland Study Area and 1996 Landslide Locations Landslide Areas Debris Flows In Valley Bottoms Steep Bluffs Along The Rivers West (Portland) Hills Silt Soils N 1996 Landslide Occurrence miles Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 13

30 Wildland Fire Multnomah County Hazard Analysis Summary of Risk Factors Severity Score High Period of occurrence: Any time, particularly summer or fall History (2) 12 Probability of event(s): Highly likely Vulnerability (5) 35 Warning time: 0 to 3 hours Maximum Threat (10) 100 Major contributor(s): Lightning or human activities resulting in fire, fuel type and condition, vegetation, and slope Probability (7) 56 Cause injuries? Yes, and risk of death Total Score 203 Potential facilities shutdown? 14 days or more Interface is used to describe areas where homes and other structures have been built on, or adjacent to, forest and range lands. It is an intermingling of built structures with natural cover at various degrees of growth and complexity. WILDLAND FIRE PROFILE Background and Local Conditions Fire is a natural part of the ecosystem and plays an important role in shaping the environment. Wildland fires generally occur heavily wooded areas but can also impact metropolitan areas; for example, these fires occur at the interface of wooded and brush areas and developed areas. These fires can be triggered by fires in the home, fires resulting from industrial activities, fires resulting from natural hazards (for example, fires associated with lightning strikes), and other events. This hazard profile considers wildland fires in the Portland study area. Historic Frequency and Probability of Occurrence Wildland fires can occur at any time of the year but are especially likely during hot, arid periods. The probability of occurrence is high, with an occurrence probable each year. Specific historic or probable frequency data are not available for the Portland wildland fire hazard. Severity The risk of impact of major wildland interface fires can be high. Wildland fire events can cause multiple deaths, completely shut down facilities, and cause more than 25 percent of affected properties to be destroyed or suffer major damage. Historic Losses and Impacts To date, there have been no major losses due to wildland fires in the Portland study area since records have been kept. Thus, while the area has been spared the impacts of fires, it is prudent to expect that such a fire represents a threat and could occur in the Portland area (Metro 1999). While no specific events have impacted Portland a number of significant wildfires occurred during 2002 and 2003 in the national forests (Deschutes and Ochoco) west and south of Portland. These necessitated road closures on Highway 20. The recent severe wildland fires and subsequent landslides in Southern California (2003) illustrate the danger that is associated with this hazard. Designated Hazard Areas Residences have long occupied the heavily forested hillsides around Portland, and the trend of locating near undeveloped land continues to increase interface areas that in turn intensify the potential impacts of wildland fires and may increase ignition sources. Structures built in interface areas may be more vulnerable to wildland fires. Approximately 30 square miles of the 145 total square miles in the Portland study area (or 20 percent) is at risk from the wildland fire hazard. Figure 3-5 shows the at-risk regions within the Portland study area. These areas are considered to be at risk because of fuel types, vegetation, and slope characteristics. The figure shows that the western and southwestern regions of the Portland study area are at greatest risk from wildland fires. 14 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

31 Figure 3-5. Areas at Risk from Wildland Fire Hazard N Portland Area = square miles Wildfire Threat Area at Risk from Wildfires = miles Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 15

32 Page intentionally left blank. 16 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

33 INVENTORY OF ASSETS 4 This section describes the inventory of assets data used for this risk assessment. Because Portland has strong, local Geographic Information System (GIS) data, local building data were collected and evaluated to support a partial Level 2 analysis. HAZUS-MH provided data was supplemented with local data for specific categories of inventory. The local inventory data collected from Metro s Regional Land Information System (RLIS) (Metro-RLIS 2002) were consistent with tax lot information and were determined to be complete and accurate for use in HAZUS-MH. Local data were used to help classify and update HAZUS-MH data for critical facilities, transportation lifeline facilities, and utility lifeline facilities. HAZUS-MH provided data were used to estimate general occupancy inventory because those data were determined to have more consistent and reliable square-foot estimates. The inventory of assets considers, What can be lost when a disaster occurs? that is, what community resources are at risk? Assets include people, buildings, transportation systems, and other valued community resources. Appendix 2 to this risk assessment report, Data Summary, provides a summary of the data collection efforts conducted for the Portland pilot project. Where available, both spatial and non-spatial local data were gathered, compiled, and analyzed. In addition to local inventory data, local GIS data also assisted Level 2 hazard modeling (as discussed in Section 5.0, Loss Estimates). Inventory data for Portland are discussed under the following categories: (1) population, (2) general building stock (aggregate inventory), and (3) critical facilities (site-specific inventory). Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 17

34 Population. The 2001 population of the City of Portland was 536,240 (Portland Development Commission 2003a). After a comparison of this total to the HAZUS-MH provided data, which estimates a population of 542,608, it was determined that HAZUS-MH population data would be used without modification. Because the two totals are similar, using the HAZUS-MH provided data would not affect the quality of the results. The HAZUS-MH estimates are slightly higher because of how the study area was defined (see Section 2.0, Identification of Hazards). Figure 4-1 shows the distribution of the general population for Portland by census tract. Figure 4-1. Distribution of General Population by Census Tract N Population Density [per square mile] 10,000 and Over 5,000 to 9,999 1,000 to 4, to 999 Less than miles In 2000, the per capita income in the Portland/Vancouver PMSA was $33,947, and the median family income for the same period was $53,700. The median household income for Multnomah County was $41,278 in 2000 (Oregon Natural Hazards Workgroup 2003). As of 2000, there were 229,333 households in the City of Portland (Metro-RLIS 2002), or an average of 2.37 individuals per household based on a population estimate of 542,608. Today, about 21 percent of Multnomah County s population is defined as minority (Oregon Natural Hazards Workgroup 2003). DMA 2000 requires that social vulnerabilities be analyzed to consider those people that may be subject to disproportionate impacts. Vulnerable populations are considered to be most susceptible to hazards based on a number of factors, including their physical and financial ability to react or respond during a hazard and the location and construction quality of their housing. Based on HAZUS-MH data, the vulnerable populations in the City of Portland evaluated for the purposes of this pilot project include the following: 18 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon

35 ? Low Income - 50,937 households with an annual income of less than $20,000 (22.2 percent of the population). Based on an average of 2.37 persons per household, this represents 120,721 low income persons.? Elderly 62,931 elderly persons (65 years or older) living in the city (11.6 percent of the total population). The low income and elderly populations are not mutually exclusive and may include individuals that appear in both categories. For the purposes of this pilot project, such overlap is not considered. Figures 4-2 and 4-3 illustrate the distribution of low income and elderly populations in the Portland study area, respectively. Figure 4-2. Distribution of Low Income Population by Census Tract Low Income Households Density Map [per square mile] 2,000 and over 1,000 to 1, to to 499 Less than 100 N miles Figure 4-3. Distribution of Elderly Population by Census Tract Elderly Population Density Map [per square mile] 2,000 and Over 1,000 to 1, to to 499 Less than 100 N miles Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon 19

36 Changes in population and development over time can increase the potential impacts associated with hazards and affect how agencies prepare for and respond to disaster. Therefore, DMA 2000 requires an analysis of development trends. Figure 4-4 illustrates the annual population change by county in the Portland/Vancouver PMSA from 1991 to The figure shows that since 1991, the regional population has exhibited a positive growth pattern. Multnomah County has grown between 5,000 and 10,000 persons annually, or about 13 percent over the 10-year period. Figure 4-4. Annual Population Change by County for the Portland/Vancouver PMSA (Metro-RLIS 2002) Figure 4-5 illustrates each county s share of the total population for the Portland/Vancouver PMSA for 1970 and The figure shows that suburban county growth in Washington, Yamhill, Clackamas, Clark, and Columbia Counties has been greater than population growth in urban, Multnomah County in the last 30 years. Therefore, Multnomah s percentage of the total regional population has decreased from 51 to 35 percent. Figure 4-5. Change in the Share of County Population for the Portland/Vancouver PMSA (Metro-RLIS 2002) 1970 Population (1,078,100) 2000 Population (1,918,009) Washington 15% Yamhill 4% Clackamas Yamhill 4% 15% Clark 12% Columbia 3% Washington 23% Clackamas 18% Clark 18% Multnomah 51% Multnomah 35% Columbia 2% 20 Risk Assessment Pilot Project Results for DMA 2000 Plan City of Portland, Oregon