HAZUSMH
A Revolution in Risk Assessment
Barbara Schauer

Introduction
Hurricanes, earthquakes, and floods annually account for billions of insured and uninsured losses in the United States. On average, two hurricanes strike the United States coastline each year. Losses from floods account for more than 75 percent of federal disaster declarations. Although earthquakes are infrequent events, they can be sudden, devastating, and costly. While many natural events are not disastrous, they still take a toll in terms of life and property, cumulatively contributing to the high cost of disasters in every state. Population growth and subsequent development in vulnerable areas, such as along U.S. coastlines, have only increased the risks.
As a direct result of the natural disasters of the early 1990s, particularly the Northridge earthquake in 1994, Hurricane Andrew in 1992, and the midwest floods in 1993, a fundamental shift in thinking occurred that focused attention on incorporating prevention and risk reduction, along with response and recovery, into emergency management. This approach to disasters includes promoting increased public awareness of the risks from natural hazards, and encouraging communities and individuals to take preventive actions. To support the need for sound risk assessment, the Federal Emergency Management Agency (FEMA) of the Department of Homeland Security (DHS) contracted with the National Institute of Building Sciences to develop a tool, implemented through a PC-based Geographic Information System (GIS), to help identify hazards and assess risks nationwide. This tool, called Hazards U.S. (HAZUS), was initially released by FEMA in 1997 and included a state-of-the-art earthquake model. A recent release is the first multihazard version, HAZUSMH which includes new flood and hurricane models. HAZUSMH allows users at all levels of government and the private sector to assess potential damage and loss from earthquakes, hurricane winds, and floods to help in determining the preventive actions needed to reduce loss of life and property. Damage and loss estimates calculated with HAZUSMH are also useful for preparing for emergency response before, during, and after natural disasters.
Research is underway to determine the usefulness of NASA-derived remote sensing data under a research contract with NASA's Earth Science Enterprise (ESE). Remote sensing data has the potential for better defining urban geographic features and providing more current information than might otherwise be found in many local communities. NASA seeks to make its data available to support improved decision making for mitigation of potential loss or response to disaster events. The benefits of the widespread use of HAZUSMH combined with high resolution data, will be improved risk assessment, improved planning and response for emergency managers, and increased awareness and implementation of prevention activities. The bottom line is: HAZUSMH is designed to help reduce the loss of lives (and injuries) and reduce economic damage as a result of natural disasters.

HAZUS Background
The first release of the HAZUS97 software occurred in 1997 following a development process that included a study of the state-of-the-art earthquake loss estimation methods, methodology and software development, and pilot testing. Three firms have participated in the development of methodologies and software: PBS&J (formerly RMS and Durham Technologies, Inc) for the earthquake model and software shell; Applied Research Associates, Inc. for the hurricane model; and ABS Consultants (formerly EQE International) for the flood model.
Major achievements during earthquake methodology development include:

n the adoption of USGS hazard maps in the hazard database;
n the use of census tracks as the lowest reference unit for a study site and the use of national databases to make the model nationally applicable;
n development of occupancy and structural building classification systems;
n the use of spectrum-capacity analysis for determining building damage instead of Modified Mercalli Intensity (MMI);
n and development of models for determining casualties, shelter requirements, debris, and an indirect economic loss.

After the initial release of HAZUS97, the model was revised three times—HAZUS99, HAZUS99-SR1, and HAZUS99-SR2 were released in 1999, 2001, and 2002, respectively. These versions not only corrected problems in HAZUS97, but also incorporated significant changes and improvements. HAZUSMH represents the last in this series of enhancements.
Development of new flood and hurricane wind models began in 1997 with a process similar to that used for the earthquake model that included a review of the state-of-the-art of loss estimation, followed by methodology and software
development, and internal testing. Major achievements in development of the flood model include:
n sophisticated capability for assessing riverine and coastal flooding;
n damage functions for all classes of buildings and essential facilities;
n estimates of damage to transportation and water utility lifelines, agricultural areas, and vehicles;
n estimates of debris quantities;
n shelter requirements;
n estimates of direct economic losses based on physical damage to structures, contents, and building interiors;
n and evaluation of the effects of flood warning and flow velocity effects.

The hurricane model is based on an existing peer-reviewed and validated hazard model that describes the entire track and wind field of a hurricane or tropical storm and has been adapted as the basis for hazard characterization in HAZUSMH. A major achievement in the hurricane model is the capability to model physical damage using engineering-based load and resistance analysis of building component performance for all classes of residential, commercial, and industrial buildings, with both wind-induced pressure and windborne debris impacts included. The model estimates losses by combining damage estimates with empirical cost estimating techniques for building repair and replacement of structures, contents, and building interiors. This method is an advancement beyond other modeling approaches that estimate potential losses based on historical financial loss records from various intensity windstorms. The load-resistance-damage-loss framework of the hurricane model allows users to reliably examine the quantitative effects of specific preventive measures by modeling building components with increased resistances. The benefits of these kinds of actions can then be quantified and compared on a regional basis for a number of building classes. The model also estimates building debris generation and post-storm shelter requirements.

Remote Sensing Data in the Hurricane Model of HAZUSMH
HAZUSMH includes the best available non-proprietary, nationwide datasets as a starting point for local analysis and are appropriate for regional assessment of the indication of risk. However, for more refined analysis for planning, enhancing, and customizing HAZUSMH with local data and local knowledge is essential. To this end, NASA-derived remote sensing data is being examined for application in HAZUSMH under an applied research contract with NASA's Earth Science Enterprise (ESE). This research includes close cooperation and participation of staff from the National Institute of Building Sciences, Applied Research Associates, Inc., and NASA researchers from the Stennis Space Center (SSC).
The current cooperative research includes the integration of NASA-developed technology into the hurricane model and will provide for its additional development of as well as promote the mainstream use of NASA products. There is currently no application that can be used for assessing risk and forecasting the potential damage and loss associated with the combined action of wind, the rise in water level (storm surge and tide), and the battering of coastal structures from hurricane-induced waves. The use of NASA remote sensing data in HAZUSMH could significantly improve both coastal flooding forecasts, due to improved digital elevation data, and provide the potential for much improved aerodynamic roughness estimation essential for the hurricane windfield model. SSC staff are supporting this research by selecting remote sensing data, with an emphasis on NASA assets, and analyzing it for its usefulness in improving the digital elevation models (DEMs) and optimizing the land cover classification scheme needed for estimating aerodynamic roughness. SSC also has the role of validating and verifying the benefits of these improvements to society. NASA's remote sensing data and analysis capabilities are essential to this application.
The following areas of study are underway:

n Integrating the WAVEWATCH III deepwater wave model, the successor of WAVEWATCH II developed at NASA, into the hurricane model.
n Examining the use of NASA's research results to develop improved digital elevation models along the coast based on remotely sensed data. Improved digital elevation models are key to improving coastal flooding predictions. The current digital elevation model (DEM) used in HAZUSMH employs 30-meter USGS data. A joint sensitivity analysis by NASA/Stennis and ARA of the higher resolution DEMs for coastal flooding will be performed to provide accuracy and precision goals for DEM development and to better understand the effects on coastal flooding loss estimation. The assumption is that high resolution DEMs are essential for modeling coastal flooding effects on barrier islands and the treatment of dune erosion during storms. The current applied research is designed to answer this and other questions.
n Researching a new concept to enable improved estimation of aerodynamic surface roughness using NASA's remote sensing data. This innovative concept for improved estimation of the aerodynamic surface roughness represents a new application of NASA remote sensing data. The potential for this totally new concept extends beyond uses in the HAZUSMH tool and includes innovations for wind tunnel testing as well as design of new buildings for wind loads.
NASA remote sensing data will also be used in the improved estimation of aerodynamic roughness, which is currently estimated in HAZUSMH using the Multi-Resolution Land Characteristics (MRLC) land use and land cover (LULC) database coupled with expert judgment. The current research will investigate a method to calculate aerodynamic roughness from remote sensing data. NASA will provide a key role in the selecting data, optimization of land cover classification scheme, processing the data, and verifying and validating.

Remote Sensing Data in the Flood Model
of HAZUSMH
The use of remote sensing data is being studied for analysis of riverine flooding. DEM data derived from airborne and spaceborne sensing systems will be evaluated to determine its utility in hydraulic and hydrologic models. Flood model users might also benefit from data obtained from NASA satellites in real-time for rapid loss estimation.
The flood model consists of two basic analytical processes: flood hazard analysis and flood loss estimation analysis. Hazard characterization in the flood model produces estimated flood depths for riverine and coastal flooding scenarios. Algorithms have been developed to define the extent of the floodplain and to interpolate flood elevations between user-supplied digital cross sections. The algorithms were developed to accommodate the most detailed digital topographic and flood elevation data available to the user while minimizing the user interaction required. The flood model accepts user-supplied data for developing flood depth and/or flood depth frequency information, and establishes the spatial distribution of that information. The result of the analysis is a Geographic Information System model in a grid format with each cell attributed with flood depth information. Flood depth is the difference between flood and ground surface elevations at each grid cell. The ground surface model is a grid-cell based digital elevation model. The flood surface model is also grid-cell based.

Conclusion
By incorporating information from remote sensing data into HAZUSMH the changes in the biophysical characteristics of regional and local communities may be factored into sophisticated risk models. The proposed methods are expected to provide a significant improvement in the quantification of coastal wind and flooding risk, provide a framework for evaluating the cost effectiveness of hurricane mitigation programs for coastal communities, and assist with real-time assessment of flooding conditions. The result will provide a much improved capability and accuracy for local communities to assess the risk and the resulting damage produced by hurricanes and floods. There are expected to be measurable economic benefits to communities and decision-makers who use these tools.

Obtaining HAZUSMH
Federal, state, and local government agencies and the private sector can now order HAZUSMH free of charge from the FEMA Distribution Center. Log onto www.fema.gov/hazus to download an order form and to learn more about HAZUSMH.

About the Author
Barbara Schauer, PE, is a Senior Project Manager with the National Institute of Building Sciences. She is managing the development of HAZUSMH under a contract with FEMA.

Photos and illustrations courtesy of FEMA.

Back