William B. Samuels, Jonathan M. Pickus, Paul Bryant, Adrian Linz

Development of an Automated Tool to Provide Logistics Support for Natural Disasters

FEMA's Response Resources and Sustainability (RRS) program provides emergency managers with answers to the following disaster relief questions: (1)what resources are needed, (2) who are the potential sources of relief supplies, and (3) where are the resources to be delivered. The RRS is a module of the Consequence Assessment Tool Set (CATS)software which estimates damage to the population and infrastructure from natural disasters such as hurricanes and earthquakes. Based on the extent of damage predicted by the CATS models, the selection of suitable mobilization sites is performed based on the following criteria: (1) distance from nearest airstrip, (2) warehouse space between 40,000 - 80,000 square feet, (3) Federal facility with physical security measures in place. Up to date coverages of vendors capable of providing disaster relief commodities such as ice, building materials, bottled water, etc. is maintained through the use of the EQUIFAX business database. These coverages are accessed by the RRS software to provide locations of local sources of relief supplies. The RRS application was written in Avenue. The CATS software is a combination of AML, Avenue, FORTRAN, C and UNIX shell scripts.

Introduction

The purpose of the Response Resources and Sustainability (RRS) program is to provide emergency planners with information about the population at risk from a disaster, the amount of relief supplies needed to support the population, the location of suitable mobilization sites for receiving supplies, and the identification of local sources of commodities.

The Federal Emergency Management Agency (FEMA) is responsible for ensuring the establishment and development of policies and programs for emergency management at the Federal, State and local levels. This responsibility includes the development of a national capability to mitigate against, prepare for, respond to and recover from the full range of emergencies, i.e., natural and technological disasters and national security emergencies. A priority of FEMA's Operations Support Directorate is to enhance the agency's response capabilities and readiness posture (FEMA, 1996a).

Based on the lessons learned from past disaster operations, the Federal Government must have the capability to provide immediate support to States whose emergency resource inventories have become quickly overwhelmed or depleted in their attempts to respond to catastrophic disasters (FEMA, 1996b). Observations from previous disaster relief support efforts include the following points: (a) deployment was slow, (b) resource movement was reactive versus proactive, and (c) multiple requests for the same disaster relief item. (MSLO, 1996)

To accomplish FEMA's emergency effort in disasters such as hurricanes, earthquakes and floods, the agency must quickly quantify relief supply requirements, rapidly identify qualified local mobilization sites to store these supplies and quickly re-supply depleted items through local vendors as needed (Vasconez, 1996). The RRS was developed to assist in meeting these needs. The RRS application enhances Esri's standard Arcview project file by providing customized menus and Avenue scripts to automate the location of mobilization sites and local vendors of relief supplies.

Methodology

The RRS application is a module derived from the CATS application and was designed to be used as a tool to support emergency management decisions. It was developed using Esri's Arcview and Avenue application code. Consequently, all user interaction and analytical results are performed through the standard functionality of Arcview, as well as the enhanced Avenue application coding.

Inputs and Outputs

Figure 1 shows a flow chart for the RRS inputs and outputs. The RSS inputs include CATS generated damage contours from either a hurricane or earthquake (CATS, 1996). The damage contours can be generated using either a deterministic (severe, moderate, light) or probabilistic (statistical distribution of severe, moderate, or light) approach. The damage estimates are produced for specific structure types (i.e., single family dwellings, mobile homes, etc.). Demographic data input is based on census block group centroids with attributes describing the population distribution by housing type (Census, 1990). Planning factors used to compute the amount of relief supplies needed, came from the Red Cross, United Nations, and U.S. Army Field Manual, as summarized in the Joint Electronic Battlebook (JEB, 1996). To locate suitable mobilization sites, four inputs are required: Federal facilities, airports, runways, and landing requirements for military transport aircraft (C-130 or C-141). Federal facility locations were provided by the GSA all agency (GSAALLST) coverage in the FEMA Master File (FEMA, 1992). Public use airports and runways were provided by the Bureau of Transportation Statistics National Transportation Atlas (BTS, 1996). Ranking factors are user specified weights (values from 1 - 10) which specify the relative importance of the criteria used to select a suitable mobilization site. These criteria are: sufficient warehouse space (40,000 - 80,000 sq. ft.), proximity to airport, and Federal facility with physical security measures in place. The RRS application produces three primary outputs in the form of maps and tables. These are: (1) the quantity of relief supplies needed to support the population, (b) the location of selected mobilization sites and airports, and (3), the location of local sources of relief supplies. The methods used to compute these outputs are described below.
Response Resources Sustainability Flow Chart

Figure 1. Flow chart for The Response Resource and Sustainability (RRS) application.

The RRS program presumes that a damage contour coverage has previously been generated through the CATS application and loaded into the RRS project file. In addition, due to the focus of the RRS module, several unique databases must be loaded as well. These include the airport database, the runway database and the GSSALLST database (Federal facilities used for mobilization sites). These databases are maintained separately to preserve them in their source format for ease of future database updates as well as generic application implementation.

RRS Modules

Calculation of Relief Support Requirements

The calculation of relief support requirements is performed by intersecting the Census block group centroids with the CATS damage contours for a specific housing type. For each damage contour, the population for a specific State (in the housing type being analyzed) is multiplied by the probability of severe, moderate, and light damage. The values of these three population risk categories (severe, moderate, light) are summed over all the damage contours to yield the total population at risk. This value is then input to a table of planning factors to yield the quantities of relief supplies: tents, cots, blankets, water, meals, etc. Figure 2 shows the results of this RRS module for estimated damage to the Gulf Coast region from Hurricane Opal. The damage contours show the probability of moderate damage to single family dwellings (SFD). For the innermost band, Prob(moderate) = 20-25%, for the outermost band, Prob(moderate) = 1-2%. The population at risk in the panhandle section of Florida was estimated to be over 90,000. Planning factors were applied to this value to yield the quantities of supplies shown in figure 2.
Calculation of Relief Supplies Needed

Figure 2. Results of the RRS module which calculates the quantity of relief supplies to support the population at risk.

Mobilization Sites

A typical RRS application requires the user to initially define the scope of the analysis through a suite of user preference parameters initiated from a drop-down menu option. The criteria describe which potential mobilization sites could be used to store critical supplies. Once defined, the user applies these preferences and the mobilization sites are automatically highlighted in the view. The user may then query a site to exhibit its descriptive information.

The user may then apply a scoring algorithm to these selected sites. A scoring algorithm can be created by the user or one of several pre- defined algorithms applied. The results from this process are comprised within a table which contain the selected mobilization sites' descriptive data and sorted on the subsequent scoring algorithm’s rank. The initial window view size is then modified to accommodate the table which is promptly displayed directly below it. At this point, the user may select one or more ranked sites in the table and its corresponding graphic icon will be highlighted within the view. This function provides a single display for the user to view both the graphic map with its associated geo-referenced features and the mobilization sites ranked attributes at the same time. In addition, graphic lines are drawn connecting mobilization sites with the airport that is closest to it.

User Preferences

Prior to defining the user preferences, the appropriate damage contour coverage must be identified. To accomplish this, the Avenue code checks for discernible coverage attributes that indicate whether its derived from probabilistic or deterministic analysis. This process is significant because distinct user preferences apply for each damage contour type. Once the damage contour is identified, the user is then required to define the scope of the analysis by responding to several input fields. These include the following: runway requirements for specific aircraft type (C-130: length = 3000 ft, width = 80 ft., C-141: length = 6000 ft, width = 98 ft., Air Mobility Command, 1996), a maximum search radius, a quantity of warehouse space in sq/ft, the minimum damage contour level to exclude sites from consideration, and scoring algorithm ranking factors. The ranking factors allow the user to prioritize warehouse space, distance from a mobilization site to an airport, and military landing rights. These factors are stored in a file that is used later during the scoring process. Figure 3 shows the drop-down user preference menu which prompts the user for inputs for selecting mobilization sites.

User preference menu

Figure 3. User preference menu to set criteria for selecting mobilization sites.

Locate Mobilization Sites

Suitable mobilization sites are located by applying the user preferences and querying the appropriate database. The first query performed derives the buffer area. This is defined as the area from the threshhold damage contour level out to the user prescribed “distance”.. This defines an exclusion area as well as a maximum distance to search for mobilization sites. Next, the aircraft type preference is used to locate all the runways meeting the length and width requirements of a C-130 or C-141 military transport. The selected runways are then queried against the buffer area to yield only qualified runways outside the damage region and within the user defined range. These runways are then associated with the airport database to derive the appropriate airport selection within the same buffered area. Next, the GSSALLST (Federal facilities) database is queried and all the potential mobilization sites that meet the minimum warehouse space specification are selected. These mobilization sites are then queried against the buffer zone to yield only those sites within the user specified range. The resulting mobilization site selection symbol legend is then updated to reflect two varied sized icons that represent selected (12pt.) and unselected (8 pt.) sites. This function was necessary because multiple GSSALLST (mobilization sites) records often shared the same geographic locations. As a result, the symbol color denoting selected sites was hidden by the symbol color of those unselected. Changing the symbol size distinguished selected from unselected sites when common locations existed. Finally, the map view is automatically altered to the maximum extent of the selected mobilization sites, thereby focusing the users view to the selected features of interest.

Selected mobilization sites

Figure 4. Results of RRS module to select suitable mobilization sites.

Apply Scoring Algorithm

If the user specified a scoring algorithm, then this file is automatically retrieved and parsed. All previously selected mobilization sites are correlated to the airport closest to it and a graphic line drawn between them in the map view. This visual link illustrates the association of each mobilization site to its closest airport. The distance between each site-airport pair is calculated and weighed according to the user specified ranking factors. Analogously, the warehouse space field within all selected mobilization sites is weighed per the scoring file. Finally, all airports that permit military landing rights are weighed. The scoring process results are automatically secured in a unique shapefile, loaded into the view and the point features symbolized. This shape file contains the following attributes: name, rank, distance from the closest airport, airport name, warehouse space and military landing rights. This data is exhibited directly under the map view. Consequently, the user can view both descriptive and graphic information simultaneously. This feature enables the user to select one or more mobilization sites from the table and observe its location highlighted in the map view. The user may at any time, modify the scoring file and re-run the process to create additional analysis. Because the results of each scoring process are inventoried within a unique shapefile, multiple scoring results may be compared concurrently.

Selected mobilization sites in New Orleans and rankings

Figure 5. Zoom in on mobilization sites in the New Orleans area. The table shows the ranking of the sites based on warehouse space, proximity to airport and security measures.

Commodities

This RRS module locates local vendors of disaster relief requirements. The data used in this module came from the Infomark Business-Facts database provided by Equifax Marketing Decision Systems (NDS, 1995). To develop the list of disaster relief suppliers, FEMA's Initial Response Resources (IRR) table of commodities (FEMA, 1996b) was examined. Standard industrial classification (SIC) codes were cross-referenced to this table. The Business-Facts database was queried by this list of SIC codes to extract the vendors of interest. Shape files were built from these extracts and added to the RRS project file. The table below shows the SIC codes used by this RRS module. In addition to these commodities, the Disaster Medical Assistance Team (DMAT, 1996) patient and drug treatment cache requirements were also addressed through the identification of local vendors with the SIC codes listed below: Based on the exclusion area and maximum search radius set in the user preference menu, local suppliers are identified in the area of interest. An example of locating ice manufacturing plants near the predicted damage area (from Hurricane Opal) is shown in Figure 6.
Selected ice plants

Figure 6. Results of RRS module to locate local suppliers of commodities. In this case, ice plants near the damage area are identified.

SUMMARY and CONCLUSIONS

The Response Resources and Sustainability (RRS) program can be used to locate suitable mobilization sites for receiving disaster relief supplies. The criteria for selecting a site are: warehouse space between 40,000 - 80,000 sq. ft., close proximity to an airport capable of landing a military transport aircraft, and a Federal facility with physical security measures in place. The RRS uses damage contours generated by the Consequence Assessment Tool Set (CATS) to quantify and locate expected damage from either hurricanes or earthquakes. Mobilization sites are selected outside of the damaged area but within a user specified search radius. These sites can also be ranked in priority order based on the relative importance of the three criteria listed above. Local vendors capable of providing disaster relief supplies are also identified and located outside the damaged area. Based on the population calculated to be at risk, the RRS program calculates the amount of resources required to support the population for a specified period of time. Planning factors provided by U.S. and International agencies are used in the calculations.

REFERENCES

Air Mobility Command, 1996. Airfield Suitability Information, Airfield Suitability Office Tools, January, 1996.

BTS, 1996. National Transportation Data Atlas, http://www.bts.gov/gis/ntatlas/ntad.htm1

CATS, 1996. Consequences Assessment Tool Set, Version 2.0, Preliminary User's Manual, US Federal Emergency Management Agency, Washington, DC, July 1996, 226 p.

Census, 1990. Summary Tape File 1, 1990 Census of Population and Housing, US Dept. of Census, Economic and Statistics Administration, Technical Documentation, May 1992.

DMAT, 1996. NDMS/DMAT Patient Treatment Cache, Draft Form DERO-004, 17p.

FEMA, 1996a., The Logistics Mission, Overview Briefing, June 26, 1996, 39p.

FEMA, 1996b., Initial Response Resources (IRR) Overview, February, 1996, 9p.

FEMA, 1992., FEMA Master File, Database Definitions, Government Functional Group, NP-SE-SW, July 1992.

JEB, 1996., Joint Electronic Battlebook (JEB) User's Manual, US Atlantic Command, July 1996.

MSLO, 1996., Improving the Federal Disaster Response and Support Process, Military Support and Liason Office, June 1996, 17p.

NDS, 1995. Infomark for Windows, Reference Guide, Version 6.2, National Decision Systems, Equifax Inc., Atlanta, GA.

Vasconez, K., 1996., FEMA to the Rescue, Logistics Spectrum, July, 1996, pages 10-14.

William B. Samuels
Senior Scientist
Science Applications International Corporation
Hazard Assessment and Simulation Division
1710 Goodridge Drive
McLean, VA 22102
Telephone: (703) 556-7074 Fax: (703) 356-8408
email: samuels@zippy.saic.com
Jonathan M. Pickus
Geographer
Science Applications International Corporation
Hazard Assessment and Simulation Division
1710 Goodridge Drive
McLean, VA 22102
Telephone: (703) 827-4814
Fax: (703) 356-8408
email: Jonathan.M.Pickus@cpqm.saic.com
Paul Bryant
Physical Scientist
Federal Emergency Management Agency
Mitigation Directorate
500 C St. SW
Washington, DC 20472
Telephone: (202) 646-3607
fax:(202) 646-4652
email: PBryant@fema.gov
Adrian Linz
Computer Specialist
Federal Emergency Management Agency
Mitigation Directorate
500 C St. SW
Washington, DC 20472
Telephone: (202) 646-3349
fax:(202) 646-4652
email: ALinz@fema.gov