Journal of Emergency Management

management with many of the characteristics of what participatory processes influencing recovery planning
Rittel and Webber74 call a “wicked problem.” Recovery progress. Although there is still much dependence on
is affected by immediate actions and choices as well as external support for recovery planning, federal recov-
by pre-existing conditions, capabilities, and a plethora ery programs present an opportunity for bottom-up
of social, institutional, and community relationships. planning.15 Challenges remain in navigating existing
This may also explain the uneven, “lurching” pattern planning structures and development pressures while
of recovery.15 The literature points to key considera- addressing risk and environmental concerns.71
tions to understand the forces influencing recovery,
including recognizing recovery as a local endeavor, It should be noted that two weeks prior to
understanding the tradeoffs and diversity of goals of Hurricane Maria, the island was struck by Irma, a
recovery, and the role of planning in recovery. This Category 5 storm that heavily damaged the island’s
study examined these aspects of recovery and has infrastructure. Since Maria, there was also a power-
highlighted important themes that emerge related ful earthquake, storms, and the pandemic that hit
to the Puerto Rico experience—equitable recovery, the island. Sea level rise, climate change, and other
leadership and participation, and the challenges of stressors will require more attention to community
recovery planning. planning and management.61 Frequent and repeated
disasters not only have both dulled efforts to prepare
In Puerto Rico, concerns regarding equity and and rebuild but also demonstrate the need for con-
fairness emerged as all too familiar issues especially tinuous improvements, training and education, and
with regard to the nature, pace, and extent of recov- broader reforms to support resilience. The research
ery and the priorities and allocation of goods and in this paper highlights the need for further training
services necessary for the rebuilding and restoration and capacity building, not only just for responders
of damaged communities. Respondents favored recov- and emergency managers but also others in the com-
ering more equitably over other recovery pathways. munity. Understanding of the different roles in recov-
Attention to the human consequences of the disaster ery41 is essential to effective targeting of training and
may have served to mortgage concerns about the envi- capacity building initiatives.
ronment and green recovery10-12,75 as well as other
societal needs.16,29,57 The recovery experience in Puerto Rico highlights
longstanding concerns as to structural inequities,
Leadership and participation are key to both colonial legacies, racism, sexism, and discriminatory
recovery and equity, with the study highlighting behaviors, which have been exacerbated during this
their influence on recovery trajectories and priorities. disaster. While some of these problems can be reduced
Disaster governance characterized by top-down or through training and education, there is also con-
bottom-up approaches underlines a choice between tinuing need for empowerment and engagement of
recovery “speed and deliberation,”6,24 a finding con- subordinated communities and mitigating corruption
firmed by survey respondents. The success of commu- before, during, and after disaster recovery. A concerted
nity-driven response in the immediate aftermath was effort to “learn from this disaster” with a willingness
highlighted, but there is room for government leader- to review and critique the plans and actions of fed-
ship to integrate community into long-term recovery eral, state, and local agencies will contribute to deeper
decisions. understanding of the theories and practices of building
back better18,20,25,26 and help to transition from short-
Recovery planning offers opportunities for com- term solutions to addressing longer-term challenges.27
munity engagement and producing collective and
equitable visions of long-term recovery, but the qual- ACKNOWLEDGMENTS
ity of participation and collaboration is driven by The research team would like to thank Dr. Lily Bui, Dr. Pradip
local planning capacity and mechanisms. This study Pant, Dr. Antares Ramos Alvarez, Dr. Maritza Barreto Orta, and
highlighted a diversity of plan existence and plan- Ernesto Diaz for their support and contributions to the team’s efforts
ning capacity among Puerto Rico municipalities, with and travels to the island. Also acknowledged is the support of the

Special Issue on Puerto Rico

Journal of Emergency Management 249
Vol. 19, No. 8

National Disaster Preparedness Training Center and the Pacific 12. Mendez-Tejeda R, Pérez-Valentín KA, Barreto-Orta M: Impact
Urban Resilience Lab at the University of Hawaii, as well as the of extreme weather events on the beaches of Puerto Rico: The case
Puerto Rico Sea Grant College Program. of Ocean Park, San Juan. Am J Mar Sci. 2020; 8(1): 1-5.
13. Castro Rivera AL, Marrero T, Ramirez DL: Temporada de
Karl Kim, PhD, Department of Urban and Regional Planning, National Huracanes 2017 en Puerto Rico: Lecciones Aprendidas. 2020.
Disaster Preparedness Training Center, University of Hawai’i at Manoa, 14. Office of Recovery, Reconstruction, and Resilience: COR3 trans-
Honolulu, Hawaii. ORCID: https://orcid.org/0000-0003-0528-8747. parency portal. Available at https://www.recovery.pr/en. Accessed
December 18, 2020.
Roberto Porro, MURP, Department of Urban and Regional Planning, 15. Garcia I, Olshansky RB, Carrasquillo D: Puerto Rico lurches
National Disaster Preparedness Training Center, University of Hawai’i toward recovery: Two years after Hurricane Maria, a bottom-up
at Manoa, Honolulu, Hawaii. ORCID: https://orcid.org/0000-0003-2403- community planning initiative is under way. 2019. Available at
1584. https://www.planning.org/planning/2019/aug/puertoricorecovery/.
Accessed December 2020.
Jiwnath Ghimire, PhD, Department of Urban and Regional Planning, 16. Berke PR, Kartez J, Wenger D: Recovery after disaster: Achieving
University of Hawai’i at Manoa, Honolulu, Hawaii. sustainable development, mitigation and equity. Disasters. 1993;
17(2): 93-109. DOI: 10.1111/j.1467-7717.1993.tb01137.x.
Lillian Ramirez Durand, MS, Puerto Rico Sea Grant College Program, 17. Rubin CB, Saperstein MD, Barbee DG: Community recovery
University of Puerto Rico, Mayaguez, Puerto Rico. from a major natural disaster. FMHI Publ. 1985; 87: 1-298.
18. Olshansky RB: How do communities recover from disaster? A
Ruperto Chaparro Serrano, MS, Puerto Rico Sea Grant College review of current knowledge and an agenda for future research. In
Program, University of Puerto Rico, Mayaguez, Puerto Rico. 46th Annu Conf Assoc Coll Sch Plan. 2005.
19. Berke P, Cooper J, Aminto M, et al.: Adaptive planning for
Bernardo Gonzalez, MURP, Department of Urban and Regional disaster recovery and resiliency: An evaluation of 87 local recovery
Planning, National Disaster Preparedness Training Center, University plans in eight states. J Am Plann Assoc. 2014; 80(4): 310-323. DOI:
of Hawai’i at Manoa, Honolulu, Hawaii. 10.1080/01944363.2014.976585.
20. Smith G, Birkland T: Building a theory of disaster recovery:
Eric Yamashita, MURP, Department of Urban and Regional Planning, Institutional dimensions. Int J Mass Emerg Disasters. 2012; 30(2):
National Disaster Preparedness Training Center, University of Hawai’i 147-170.
at Manoa, Honolulu, Hawaii. 21. Smith GP, Wenger D: Sustainable disaster recovery: Operation-
alizing an existing agenda. In Handbook of Disaster Research. New
REFERENCES York: Springer, 2007: 234-257. DOI: 10.1007/978-0-387-32353-4_14.
1. National Hurricane Center: Tropical cyclone report: Hurricane 22. Government of Puerto Rico: Código Municipal de Puerto Rico,
Maria. 2019. 2020: 525.
2. Central Office for Recovery, Reconstruction and Resiliency 23. Federal Emergency Management Agency (FEMA): Hurricane
(COR3): Transformation and innovation in the wake of devastation: season FEMA after-action report. 2017.
An economic and disaster recovery plan for Puerto Rico. 2018. 24. Platt S, So E: Speed or deliberation: A comparison of post-dis-
3. George Washington University: Ascertainment of the estimated aster recovery in Japan, Turkey, and Chile. Disasters. 2017; 41(4):
excess mortality from Hurricane Maria in Puerto Rico. 2018. 696-727. DOI: 10.1111/disa.12219.
Available at https://publichealth.gwu.edu/sites/default/files/down- 25. Kim K, Olshansky RB: The theory and practice of build-
loads/projects/Prstudy/Acertainment. ing back better. J Am Plann Assoc. 2014; 80(4): 289-292. DOI:
4. National Centers for Environmental Information (NCEI): 10.1080/01944363.2014.988597.
Costliest US tropical cyclones. 2020. 26. Fernandez G, Ahmed I: Build back better approach to disaster
5. Johnson LA, Hayashi H: Synthesis efforts in disaster recovery recovery: Research trends since 2006. Prog Disaster Sci. 2019;
research. Int J Mass Emerg Disasters. 2012; 30(2): 212-238. 1:100003. DOI: 10.1016/j.pdisas.2019.100003.
6. Olshansky RB, Hopkins LD, Johnson LA: Disaster and recovery: 27. Finucane ML, Acosta J, Wicker A, et al.: Short-term solutions
Processes compressed in time. Nat Hazards Rev. 2012; 13(3): 173- to a long-term challenge: Rethinking disaster recovery planning
178. DOI: 10.1061/(ASCE)NH.1527-6996.0000077. to reduce vulnerabilities and inequities. Int J Environ Res Public
7. Kim K, Bui L: Learning from Hurricane Maria: Island ports Health. 2020; 17(2): 482. DOI: 10.3390/ijerph17020482.
and supply chain resilience. Int J Disaster Risk Reduct. 2019; 39: 28. Kim K, Foley D: Report to kettering foundation, disasters and
101244. DOI: 10.1016/j.ijdrr.2019.101244. democracy project. 2014.
8. Farber DA: Response and recovery after Maria: Lessons for dis- 29. Mehta A, Brennan M, Steil J: Affordable housing, dis-
aster law and policy. Rev Jurisdica Univ P R. 2018; 87(3): 743-771. asters, and social equity: LIHTC as a tool for preparedness
DOI: 10.2139/ssrn.3174466. and recovery. J Am Plann Assoc. 2020; 86(1): 75-88. DOI:
9. Hamdani K, Kramer Mills C, Silva J, et al.: Small businesses 10.1080/01944363.2019.1667261.
and the 2017 hurricanes. 2018. Available at https://www.new 30. Meerow S, Pajouhesh P, Miller TR: Social equity in urban
yorkfed.org/medialibrary/media/outreach-and-education/24idwes- resilience planning. Local Environ. 2019; 24(9): 793-808. DOI:
rico/2018/PRSBS-2018-Hurricanes.pdf. 10.1080/13549839.2019.1645103.
10. Reardon S: Hurricane Maria’s wrath leaves clues to coral reefs’ 31. March A, Kornakova M (eds.): Urban Planning for Disaster
future. Nature. 2018; 560(7719): 421-423. Recovery. New York: Butterworth-Heinemann, 2017: 31.
11. Barreto-Orta M, Méndez-Tejeda R, Rodríguez E, et al.: State 32. Domingue SJ, Emrich CT: Social vulnerability and procedural
of the beaches in Puerto Rico after Hurricane Maria (2017). Shore equity: Exploring the distribution of disaster aid across counties in
Beach. 2019; 87(1): 8. the United States. Am Rev Public Adm. 2019; 49(8): 897-913. DOI:
10.1177/0275074019856122.

Special Issue on Puerto Rico

250 Journal of Emergency Management
Vol. 19, No. 8

33. Emrich CT, Tate E, Larson SE, et al.: Measuring social equity in 53. Siders AR: Social justice implications of US managed retreat
flood recovery funding. Environ Hazards. 2020; 19(3): 228-250. DOI: buyout programs. Clim Change. 2019; 152(2): 239-257. DOI:
10.1080/17477891.2019.1675578. 10.1007/s10584-018-2272-5.
34. Reardon KM: Equity planning in Post-Hurricane Katrina New 54. Muñoz C, Tate E: Unequal recovery? Federal resource distribu-
Orleans: Lessons from the ninth ward. Cityscape. 2008; 10(3): 21. tion after a Midwest flood disaster. Int J Environ Res Public Health.
35. Ehrenfeucht R, Nelson M: Planning, population loss and equity 2016; 13(5): 507. DOI: 10.3390/ijerph13050507.
in New Orleans after Hurricane Katrina. Plan Pract Res. 2011; 55. Willison CE, Singer PM, Creary MS, et al.: Quantifying ineq-
26(2): 129-146. DOI: 10.1080/02697459.2011.560457. uities in US federal response to hurricane disaster in Texas and
36. Rubin CB, Barbee DG: Disaster recovery and hazard mitigation: Florida compared with Puerto Rico. BMJ Glob Health. 2019; 4(1):
Bridging the intergovernmental gap. Public Adm Rev. 1985; 45: 57. E001191. DOI: 10.1136/bmjgh-2018-001191.
DOI: 10.2307/3134998. 56. Basta C: From justice in planning toward planning for justice: A
37. Delgado CM: Plans require participation: Disaster recovery capability approach. Plan Theory. 2016; 15(2): 190-212.
plans are ineffective without people. J Emerg Med. 2007; 5(5): 25. 57. Welsh MG, Esnard A-M: Closing gaps in local housing recovery
DOI: 10.5055/jem.2007.0021. planning for disadvantaged displaced households. Cityscape. 2009;
38. Pearce L: Disaster management and community planning, and 11(3): 19.
public participation: How to achieve sustainable hazard mitigation. 58. McEntire DA: Tenets of vulnerability: An assessment of a
Nat Hazards. 2003; 28(2-3): 211-228. fundamental disaster concept. J Emerg Med. 2004; 2(2): 23. DOI:
39. Djalante R, Holley C, Thomalla F: Adaptive governance and 10.5055/jem.2004.0020.
managing resilience to natural hazards. Int J Disaster Risk Sci. 59. Sotolongo M, Kuhl L, Baker SH: Using environmental justice to
2011; 2(4): 1-14. DOI: 10.1007/s13753-011-0015-6. inform disaster recovery: Vulnerability and electricity restoration
40. National Centers for Environmental Information: Billion-dollar in Puerto Rico. Environ Sci Policy. 2021; 122: 59-71. DOI: 10.1016/j.
weather and climate disasters: Overview. 2020. Available at https:// envsci.2021.04.004.
www.ncdc.noaa.gov/billions/. Accessed December 20, 2020. 60. Federal Emergency Management Agency (FEMA): National
41. Keenan JM: Types and forms of resilience in local planning in Disaster Recovery Framework. 2nd ed. Washington, DC: FEMA,
the US: Who does what? Environ Sci Policy. 2018; 88: 116-123. DOI: 2016.
10.1016/j.envsci.2018.06.015. 61. Santiago L, Flores D, Hong C-Y: The impact of extreme weather
42. Reece JW: In pursuit of a twenty-first century just city: The events on community risk planning and management: The case of San
evolution of equity planning theory and practice. J Plan Lit. 2018; Juan, Puerto Rico after Hurricane Maria. Urbe Rev Bras Gest Urbana.
33(3): 299-309. DOI: 10.1177/0885412218754519. 2020; 12: E20190062. DOI: 10.1590/2175-3369.012.e20190062.
43. Yiftachel O: Planning and social control: Exploring the dark 62. Neal DM: Transition from response to recovery after the
side. J Plan Lit. 1998; 12(4): 395-406. DOI: 10.1177/0885412298012 Lancaster, TX, tornado: An empirical description. J Emerg Med.
00401. 2004; 2(1): 47. DOI: 10.5055/jem.2004.0011.
44. Cutter SL, Finch C: Temporal and spatial changes in social 63. Gonzalez B, Jr, Weygan-Hildebrand C, Winchester C, et al.:
vulnerability to natural hazards. Proc Natl Acad Sci USA. 2008; Island community resilience initiative report: Incorporating lessons
105(7): 2301-2306. from Hurricane Maria into Oahu’s resilience strategy. 2018.
45. Tierney K, Oliver-Smith A: Social dimensions of disaster recov- 64. Sandler D, Smith G: Assessing the quality of state disaster
ery. Int J Mass Emerg Disasters. 2012; 30(2): 1-23. recovery plans: Implications for policy and practice. J Emerg Med.
46. Comfort LK, Birkland TA, Cigler BA, et al.: Retrospectives 2017; 11(4): 281. DOI: 10.5055/jem.2013.0144.
and prospectives on Hurricane Katrina: Five years and count- 65. Government Accountability Office: Puerto Rico hurricanes:
ing. Public Adm Rev. 2010; 70(5): 669-678. DOI: 10.1111/j.1540- Status of FEMA funding, oversight and recovery challenges. 2019.
6210.2010.02194.x. Available at http://cedd.pr.gov/Mitigacion/. Accessed December 29,
47. Kates RW, Colten CE, Laska S, et al.: Reconstruction of New 2020.
Orleans after Hurricane Katrina: A research perspective. Proc 66. Burger N, Cook CR, Finucane ML, et al.: Developing Recovery
Natl Acad Sci USA. 2006; 103(40): 14653-14660. DOI: 10.1073/ Options for Puerto Rico’s Economic and Disaster Recovery Plan:
pnas.0605726103. Process and Methodology. Geneva: UNDRR, 2020: 81.
48. Billings SB, Gallagher EA, Ricketts L: Let the rich be flooded: 67. Kennedy M, Metz D, Dezenski E, et al.: Assessing the Cost of
The distribution of financial aid and distress after Hurricane Disaster Recovery and Identifying Funding Sources in the HSOAC
Harvey. J Financial Econ. 2019; 59:1-48. Puerto Rico Economic and Disaster Recovery Plan Project. Geneva:
49. Howell J, Elliott JR. Damages done: The longitudinal impacts RAND Corporation, 2020. DOI: 10.7249/RR2861.
of natural hazards on wealth inequality in the United States. Soc 68. Nunez-Neto B, Lauland A, Aguirre J, et al.: Municipalities on
Probl. 2019; 66(3): 448-467. DOI: 10.1093/socpro/spy016. the Front Lines of Puerto Rico’s Recovery: Assessing Damage, Needs,
50. US Census Bureau: QuickFacts Puerto Rico. American and Opportunities for Recovery after Hurricane Maria. Geneva:
Community Survey 2019. 2020. Available at https://www.cen RAND Corporation, 2020. DOI: 10.7249/RR2604.
sus.gov/quickfacts/fact/table/PR/INC910219#INC910219. Accessed 69. Puerto Rico Department of Housing: Municipal recovery plan-
December 20, 2020. ning program. Municipal Recovery-CDBG. 2020. Available at
51. Bach T: The 10 US cities with the largest income inequality https://cdbg-dr.pr.gov/en/municipal-recovery/. Accessed December
gaps. US news and world report. 2020. Available at https://www. 27, 2020.
usnews.com/news/cities/articles/2020-09-21/us-cities-with-the-big 70. Junta de Planificacion de Puerto Rico: Planes de mitigacion con-
gest-income-inequality-gaps. Accessed December 21, 2020. tra peligros Naturales-Inicio. Planes de mitigación contra peligros
52. Elliott JR, Brown PL, Loughran K: Racial inequities in the naturales. Available at http://cedd.pr.gov/Mitigacion/. Accessed
federal buyout of flood-prone homes: A nationwide assessment of December 2020.
environmental adaptation. Socius. 2020; 6: 237802312090543. DOI: 71. Chaparro R: Ponencia del Programa Sea Grant de la Universidad
10.1177/2378023120905439. de Puerto Rico sobre el propuesto Reglamento Conjunto 2020 para

Special Issue on Puerto Rico

Journal of Emergency Management 251
Vol. 19, No. 8

la Evaluacion y Expedicion de Permisos Relacionados al Desarrollo, 74. Rittel HWJ, Webber MM: Dilemmas in a general theory of plan-
Uso de Terrenos y Operacion de Negocios (RC 2020). 2020. ning. Policy Sci. 1973; 4(2): 155-169.
72. Schwab J, Topping K, Eadie C, et al.: Planning for Post-Disaster 75. Miller S: Greenspace after a disaster: The need to close the gap
Recovery and Reconstruction. Chicago, IL: American Planning with recovery for greater resilience. J Am Plann Assoc. 2020; 86(3):
Association, 1998. 339-348. DOI: 10.1080/01944363.2020.1730223.
73. Anguelovski I, Shi L, Chu E, et al.: Equity impacts of urban land 76. Federal Emergency Management Agency (FEMA): Mitigation
use planning for climate adaptation: Critical perspectives from the Assessment Team Report: Hurricanes Irma and Maria in Puerto
global North and South. J Plan Educ Res. 2016; 36(3): 333-348. Rico. 2018.
DOI: 10.1177/0739456X16645166.

APPENDIX A: SURVEY QUESTIONNAIRE

Recovery planning survey for municipalities in Puerto Rico
  1.  What is the name of your municipality?
   a. ____________________________
  2.  Which individual(s) in your office is/are responsible for recovery planning?
   a. ____________________________
  3.  What community groups or organizations (outside of government) are involved in recovery planning?
   a. ____________________________
 4.  What funded recovery, mitigation, or adaptation programs are active in your municipality?
   a. ____________________________
 5.  What non-funded recovery, mitigation, or adaptation programs are active in your municipality?
   a. ______________________________
 6.  What funded recovery, mitigation, or adaptation programs are active in your community?
   a. ____________________________
 7.  What non-funded recovery, mitigation, or adaptation programs are active in your community?
  8.  From a scale of 1 to 10, with 1 being the least severe and 10 being the most severe, how would you rate

the extent of damage to your municipality from Hurricane Maria?
   1 2 3 4 5 6 7 8 9 10
  9.  How would you rate the status of recovery in your municipality?
   a. Not started
   b. In progress/ongoing
   c. Complete
10.  Does your municipality have a master plan?
   a. Yes
   b. No
11.  Does your municipality have an emergency plan?
   a. Yes
   b. No
    c.  No, but emergency planning is embedded in our master plan
12.  Does your municipality have a resilience plan?
   a. Yes
   b. No
    c.  No, but resilience is embedded in our master plan
    d.  No, but resilience is embedded in our emergency plan

Special Issue on Puerto Rico

252 Journal of Emergency Management
Vol. 19, No. 8

13.  Does your municipality have a standalone recovery plan?
   a. Yes
   b. No
    c.  No, but recovery is embedded in our emergency plan
    d.  No, but recovery is embedded in our master plan
14.  Please rank in order of priority what your municipality’s recovery goals are:
   a. Recover faster
   b. Recover more strongly
    c.  Recover greener (more sustainably)
    d.  Recover more equitably (taking into account whole communities and vulnerable populations)
15.  Are there good examples of municipal or community recovery projects that you believe exhibit these

priorities?
   a. Yes
    b.  If yes, which ones? _____________________
   c. No
   d. I don’t know/unsure
16.  Do these recovery projects include participatory processes?
   a. Yes
   b. No
17.  Have you ever attended a National Disaster Preparedness Training Center training?
   a. Yes
   b. No
18.  Please rank in order of priority your municipality’s training needs:
   a. Disaster recovery planning
   b. Climate change adaptation
   c. Community resilience
   d. Hurricane awareness
    e.  Coastal hazards assessment and planning
    f.  Disaster awareness for community leaders
   g. Tsunami awareness
    h.  Disaster debris management planning
   i. Small business continuity
   j. Other: ___________________
19.  May we contact you for follow-up questions?
   a. Yes
   b. No
20.  If you answered yes to #17, please fill out the following:
   Name:________________________
    Telephone: ____________________
   E-mail: ________________________

Special Issue on Puerto Rico

Journal of Emergency Management 253
Vol. 19, No. 8

Notes

Special Issue on Puerto Rico
254 Journal of Emergency Management

Vol. 19, No. 8

JEM Principles for data management, visualization, and
communication to improve disaster response management:

Lessons from the Hurricane Maria response mission

Marin M. Kress, PhD
Katherine F. Chambers, MS
Darixa D. Hernández-Abrams, MS

S. Kyle McKay, PhD, PE

ABSTRACT Key words: Puerto Rico, temporary roofing, power
restoration, geospatial analysis, Task Force Power,
Data visualization and communication are impor- Operation Blue Roof, Hurricane Maria
tant components in disaster response management.
Data management should be a basic part of emer- INTRODUCTION
gency preparation in the same way as prepositioning
essential supplies. For this preparation to be effective, Natural disasters create complex problems due
well-conceived data structures and data collection to their impacts, cascading effects, and the capac-
systems must be in place before disasters happen, and ity of coupled social-ecological systems to antici-
required hardware should be designed to operate in pate, respond, and recover from disasters.1 Response
contingency environments. However, due to challenges to natural disasters is becoming more challenging
in disaster complexities and data management, there due to the increased intensity and magnitude of
is still a pressing need for improvement. This paper extreme weather events,2 urbanization and popula-
identifies key principles to assist practitioners and tion growth,3 and the growing societal connectedness
software developers in designing and implementing brought by globalization.4 While the magnitude of the
data collection and reporting systems that can be used challenge is large, advances in technology and lessons
for data visualization during a disaster response. learned from past natural disasters have helped pro-
The authors reviewed existing literature on data and tect human lives by improving disaster management,
disaster management and incorporated their personal minimizing risks to hazardous events, and maximiz-
experiences as first responders with the US Army ing recovery efficiency.
Corps of Engineers Hurricane Maria response mission
to develop principles for improving data management A crucial component in the effectiveness of pre-
and visualization during a disaster response. These paredness, response, and recovery of disasters is data
principles are illustrated by two case studies from the visualization and communication it facilitates. Data
Task Force Power and Operation Blue Roof mission visualization is, fundamentally, “the process of produc-
efforts in Puerto Rico during 2017-2018. Suggested ing visual representations of data and the outputs of
principles include considering data management as that work”5 and can be a powerful tool that aids in dis-
part of disaster preparedness, having flexible data aster management because it “aims to enhance one’s
tools resilient to unprecedented disaster outcomes, ability to carry out a task by encoding often highly
eg, interruption of telecommunications networks, and abstract information into a visual form.”5 This visual
using diverse graphics and tools that are appropriate form allows disaster management professionals to
to their communication purpose and audience. track performance concerns and make informed deci-
sions to improve response and recovery. A key point is

DOI:10.5055/jem.0658 Special Issue on Puerto Rico
Journal of Emergency Management
Vol. 19, No. 8 255

that data visualization will not replace the judgment damage to every municipality. The combination of
required by disaster responders, but it can support storm surge inundation (up to 9 ft.), high winds, and
these dedicated professionals in stressful situations by rainfall (up to 38 in.) caused record breaking flood-
clearly and quickly providing organized information ing, mud slides, and property damage. Maria was an
needed to carry out their mission. Harnessing power- extremely destructive hurricane causing at least 65
ful new datasets for disaster management has been a deaths9,11,12 and approximately 90 billion dollars in
new focus area of research, and many new tools and estimated damages across Puerto Rico and the US
technologies have been identified and employed dur- Virgin islands.9,13 These massive human, ecological,
ing response efforts.6 However, multiple challenges and economic impacts highlight the importance—and
exist in the process of collecting, extracting, process- challenge—of accurate data collection before, during,
ing, and communicating data related to disaster man- and after a crisis.
agement. These challenges have been summarized by
others7,8 as falling into four main types: During natural disasters and other emergencies,
the US Army Corps of Engineers (USACE) can be
1.  Structural and semantical heterogene- involved in response and recovery activities through
ity of input data sources and the diverse its own authority (Public Law 84-99) or in sup-
needs of consumers. port of the Federal Emergency Management Agency
(FEMA) as the lead agency for the “Public Works and
2.  The need for ad-hoc data collection and Engineering” Emergency Support Function.14 These
modeling due to the diversity and com- support functions can include providing temporary
plexities of disasters themselves. power, temporary roofing, emergency water assis-
tance, debris management, and other types of techni-
3.  Time sensitivity of data exchange and cal assistance.14 In response to the 2017 hurricanes,
the wide array of options available to FEMA tasked USACE with providing debris manage-
illustrate data, eg, maps, graphs, and text ment, infrastructure assessments, temporary roofing,
summaries. temporary power, and the additional and unprece-
dented role of coordinating and directly assisting with
4.  Data quality and reliability of data power grid restoration in Puerto Rico.15
sources.
This paper builds on the wide-ranging and large
The unprecedented situation in the aftermath of USACE emergency response effort to synthesize a set
Hurricane Maria illustrated the reality of these chal- of recommendations, or principles, for data visualiza-
lenges. Hurricane Maria (Category 4) made landfall tion during disaster response. These principles for
in Puerto Rico on September 20, 2017, causing cata- data visualization are illustrated by two case stud-
strophic and historical damages.9 It was the strongest ies that capture the federal response to a hurricane
hurricane to make landfall in Puerto Rico since San that caused both extensive network and patch-style
Felipe II (Category 5) 89 years ago. Only 2 weeks impacts across the island of Puerto Rico: Task Force
prior, Hurricane Irma had passed near Puerto Rico, Power Restoration (TFP) and Operation Blue Roof.
and the combination of tropical-storm force winds, The TFP mission was tasked with managing recon-
heavy rainfall, and storm surge flooding in coastal struction of a geographically expansive electrical gen-
areas caused widespread power outages, damage to eration and distribution network, while the Operation
structures, uprooted trees, and at least three deaths.10 Blue Roof mission was tasked with providing assis-
When Hurricane Maria arrived, it crossed the main tance to (primarily) individual homeowners through
island of Puerto Rico diagonally from the south- the installation of temporary roofing materials. Our
east to the northwest for over 8 hours and caused post hoc analyses of Hurricane Maria are, impor-
tantly, not intended as a critique of agency response
and recovery, but instead benefit from the advantage

Special Issue on Puerto Rico

256 Journal of Emergency Management
Vol. 19, No. 8

of hindsight to look to future use of visualization and flexible enough to accommodate many scenarios. In
to suggest specific technical preparations to improve short, tools should improvise alongside the mission.
disaster response management. This principle is illustrated by the powerful (but
drastically different) situational awareness methods
METHODS that were utilized during TFP’s mission. The power
restoration mission was the first of its kind by the
Identifying best practices for disaster data USACE, and the existing data templates, data man-
management and visualization agement structures, and tools for collecting data in
the field were overwhelmed by the scale of the mis-
Three of the authors (MMK, KFC, and SKM) were sion, its unique requirements, and by the lack of cel-
directly involved in response activities of the USACE lular and internet connectivity during the first few
after Hurricane Maria, thus drew from personal expe- months of the effort. At the start of the mission, the
riences with on-site data management and visualiza- newly established TFP had to move quickly to identify
tion challenges. Collectively, these personal experi- the status of the system, its redundancies, key public
ences included the following: (1) a deployment to the service nodes, and critical population centers. Access
TFP as a Data Manager at the operations center; (2) a to detailed maps of the system was slow to obtain
deployment with Operation Blue Roof as a Quality as Puerto Rico Electric Power Authority (PREPA)
Assurance Inspector performing on-site assessments employees were unable to return to work and access
at residential structures; and (3) an extended deploy- their computers and database storage systems. The
ment with Operation Blue Roof as a Data Manager at initial map of the power system for the entire mis-
the main field operations office with responsibilities sion was a “ball and stick” representation of the 230
spanning data communication, process improvement, KV and 115 KV transmission lines and power dis-
and error detection. These deployments provided a tribution centers (Figure 1). The operational status
unique experience with a variety of data management of each line was manually updated within Microsoft
practices as they evolved during the Hurricane Maria PowerPoint to convey changes. This method was
response. In addition to the authors’ personal experi- labor intensive, but the TFP ball-and-stick map was
ence, peer-reviewed literature was surveyed at the incredibly useful in orienting decision makers to the
intersection of disaster management and visual ana- structure of the electrical grid and allowing them to
lytics, data management, and information systems, strategize the location of USACE resources to coordi-
eg, NRC 2007.8 The four principles described in this nate repair and reconstruction activities and to focus
paper and the specific recommended practices are a on where to prioritize efforts. The map also became
synthesis of field experience and academic theory. key to communication between the Unified Command
Leadership (USACE, FEMA, Emergency Restoration
We present two case study examples drawn from Coordinator, and PREPA), the Governor of Puerto
personal experiences during the Hurricane Maria Rico, and executive leadership in Washington DC.
response and then a resulting set of principles along- It can be found within many public briefing materi-
side key peer-reviewed literature resources. These als that were produced during the first month of the
principles are not intended as a comprehensive mission (October to December 2017) as well as subse-
review of data-driven disaster management,8,16 infor- quent communication materials.22
mation systems,17,18 or visual analytics.5,19-21 Instead,
we provide disaster managers and those who design The mission as directed by FEMA evolved to
data collection systems, with a plain language trans- include not just the restoration of the transmission
lation of key data visualization principles to acceler- lines, but the comprehensive restoration of overall
ate and facilitate better response in future. power generation, transmission, and distribution sys-
tems down to the last mile (consumer and residential
Case study: Task Force Power Restoration consumers). The need to have consistent, reliable,
In disaster management, often the only certainty

is uncertainty, and decision-making tools need to be

Special Issue on Puerto Rico

Journal of Emergency Management 257
Vol. 19, No. 8

Figure 1.  Electrical grid “ball and stick” representation, showing system redundancy and population centers
across Puerto Rico.22

and up-to-date information on the entire system grew transition in the functions of supporting infrastruc-
as the TFP leadership experienced a large increase ture without losing fidelity. (4) Data communication
in requests for status reports from USACE and needs within a single mission will vary, eg, status
FEMA leadership and national media. In addition, reporting, prioritization for directing future resources.
TFP volunteers would consistently rotate in and out One dashboard will not satisfy every need.
of the mission (often on 30-day deployments) and
need to be quickly brought up to speed on current Case study: Operation Blue Roof
procedures when they arrived. By December 2017, Disaster management often requires juggling
internet connectivity was stable, and ArcGIS analyst
teams embedded at the TFP headquarters were able multiple objectives simultaneously, even within a
to deploy a customizable online portal to convey this single mission. The USACE maintains the Operation
necessary information (Figure 2). This portal was Blue Roof mission, whose purpose is to provide
very well received, and the situational briefs quickly homeowners in disaster areas with fiber-reinforced
moved from paragraph and bulleted text updates to plastic sheeting (colloquially known as “blue roofs”)
seamlessly generated graphs based on daily input until permanent roofing repairs can be made.23 The
from the field. overarching mission objective is to provide temporary
roofing as quickly as possible to prevent further dam-
Overall findings from the TFP data management age to a home, thus allow people to shelter in their
experience are as follows. (1) Tools and datasets for own home. The scale of assistance requested across
fast collaboration and communication are only as use- Puerto Rico after Hurricane Maria was the third larg-
ful as the availability of the communication networks est in the history of the Blue Roof program, which had
that support them. (2) When those networks are not already been active during the 2017 hurricane season
available, first responders will resort to whatever in Florida and the US Virgin Islands in response to
existing tools they have on hand (including pen and the storms preceding Hurricane Maria.23 The chal-
paper, PDF, word documents, cell phone notes, etc.). lenge of accurately tracking thousands of cases was
(3) Tools for disaster management need to imple- compounded by power outages and limited cellular or
ment improvizational tactics by accommodating for a internet connectivity in many places. Eventually, over

Special Issue on Puerto Rico

258 Journal of Emergency Management
Vol. 19, No. 8

Figure 2.  Screenshot of the USACE TFP web viewer. Image: USACE.

Figure 3.  Idealized process for a structure participating in the Blue Roof program, with eight important steps.
ROE: Right of Entry form (authorization form signed by homeowner allowing government officials and contrac-
tors to access the property).

60,000 blue roofs were installed throughout Puerto assessment, roof repair work, and final inspection. For
Rico across all 78 municipalities.23 any data management system designed to track con-
struction progression, we strongly recommend that
As with any construction-based activity, the pro- each step change event has a date-stamp associated
gress of a single home awaiting assistance from the with it. This would allow mission staff to see not just
Operation Blue Roof mission went through multiple where in the process a property is at a single moment,
phases. Figure 3 shows an idealized progression of a but how long it took to move through each step.
single property, from the initial creation of a Right of
Entry (ROE) form signed by the homeowner, allow- This idealized progression shown in Figure 3
ing access to the property by government inspectors could be slowed for a variety of reasons, examples of
and contractors, through Blue Roof installation on which are shown in Figure 4. Each problem, whether
the structure, to the final step of payment by the incorrect location information or construction crew
government to the construction contractor. This ide- safety concerns, had to be resolved before the pro-
alized process includes eight steps, which required cess could resume. For example, there were many
at least three on-site visits to a property for initial instances of incomplete address information on an

Special Issue on Puerto Rico

Journal of Emergency Management 259
Vol. 19, No. 8

Figure 4.  Blue Roof assistance progress with example problems that could delay delivery. ROE: Right of Entry
form (authorization form signed by homeowner allowing government officials and contractors to access the
property).

Figure 5. Example of pixel graph-style visualization to display progress tracking for hypothetical Blue Roof
cases.

ROE form. Resolving this common problem required questions from mayors and community leaders while
time and resulted in the practice of latitude and lon- still respecting individual privacy, and mission lead-
gitude information being added as a standard piece of ership needed island-wide summary data to track
address information even though ROE forms were not progress and execute management actions. These
designed to capture this piece of information. data needs were in addition to the daily data needs
of the mission staff distributed across the island in
Tracking of each case at each step was essential field offices. Field office managers needed to dispatch
for conveying mission progression at different levels staff to different site visits every day, interface with
of aggregation. For example, individual homeowners contractors and suppliers, and respond to unique
who called the Blue Roof helpline needed information information requests. Figure 5 shows a pixel graph
about their case, local government liaisons needed displaying a sample timeline of 10 hypothetical cases
information about specific municipalities to answer

Special Issue on Puerto Rico

260 Journal of Emergency Management
Vol. 19, No. 8

at irregular time intervals. An ideal progression presented for multiple audiences, emphasizing differ-
would move from light blue (assessment) to dark blue ent narratives based on information needs. The alter-
(installation and inspection), then to green (paid), but native presentations of data in Figure 6 are examples
other states were possible. Pixel graphs are especially from the USACE briefing materials as of December
useful for large data sets because the human eye is 20, 2017 as a demonstration of interim status report-
drawn to colors and patterns. For example, a sud- ing 3 months after the Hurricane Maria. The plots in
den increase or cluster in inspection failures (red) Figure 6 are not outputs from the mission database
could indicate the need for additional training or the as it did not have that capability, but similar graphs
discovery of a manufacturing defect in construction were developed independently in response to mission
materials. The key function would be to present and needs. Mission leaders required information about
preserve a history of each case progression, with the two primary response tasks: field assessments of
ability to group records by different attributes or homes and installation of blue roofs. Figure 6A shows
subobjectives. a simple line plot of these data relative to calendar
date, but visual references have been added to con-
The Blue Roof mission provides a useful dem- textualize decision-making. A horizontal line shows
onstration of how a single data set may need to be

Figure 6. Alternative visual summaries of temporary roofing status during Hurricane Maria. (A) Island-wide
blue roof summary with embedded reference points describing citizen effects. (B) Island-wide blue roof sum-
mary using color to emphasize status and trends.

Special Issue on Puerto Rico

Journal of Emergency Management 261
Vol. 19, No. 8

the predicted number of blue roofs based on prior hur- communication methods. Although this paper focuses
ricanes, and vertical lines show key outcomes relative on data communication within a response effort, an
to citizens’ experiences such as milestones related to important component of data communication is shar-
household damages and psychological responses to ing an information with outside entities such as news
displacement from other hurricanes. Figure 6B pre- organizations. Groups in charge of data collection,
sents the same data using color and plain language processing, and visualization must be able to brief
to emphasize outcomes in plain language and evoke information clearly to agency leaders and the press
a more personal response. Additionally, calendar that informs the public. The design of communica-
days were replaced with time since the hurricane to tion products destined for public consumption should
reduce memory required by users, ie, “Was the storm include consultation with experienced communication
on September 15th or 25th?” These graphics demon- professionals to ensure clarity and consistency with
strate the value that could be added from alternative other communication products.
visualizations as well as the need to help users inter-
pret and contextualize data by providing visual cues. Data-collection systems that are likely to be used
by volunteers or new staff (or where there will be fre-
Overall findings from the Operation Blue Roof data quent staff turnover such as during a response effort
management experience are as follows. (1) Visualization lasting more than 1 month) should be designed to
may need to be adapted for different purposes or audi- minimize potential sources of error as much as pos-
ences. (2) The same data visualized multiple ways may sible. To avoid typographical or transcriptional errors
lead to different observations. (3) Plain language in with geospatial information, such as addresses, there
visuals can potentially increase the breadth of audi- should be as much prepopulated information as pos-
ence understanding and reduce miscommunications. sible in any data collection system. For example, most
(4) Simple changes to figures can make them more acces- parts of the United States have well-defined spatial
sible to decision-makers and nontechnical audiences. units, such as towns, counties, municipalities, or ZIP
code routes from the US Postal Service. On the whole,
Results and principles these do not change very much year to year and could
Here, we outline principles for effectively using be brought into (or activated within) data collection
systems on a state-by-state basis depending on the
data visualization and communication to inform dis- extent of the disaster. Data-capture systems that will
aster response and recovery. be used in dynamic environments where the risk of
fatigue is high should minimize the burden of data
Principle 1: preparedness is more than people and entry as much as possible. To support this user-centric
supplies: it must include data collection systems design, data-capture systems should auto-populate
designed for a full suite of mission needs and error-check fields as much as possible. To con-
tinue with the ZIP code example, typical address cap-
Resilience literature describes a cycle with the ture follows the order of (1) house number, (2) street
phases of prepare, absorb, recover, and adapt.24-27 name, (3) city, (4) state, and then (5) ZIP code. A
Often, the “prepare” phase focuses on the preposition- system with built in error-checking could start with
ing of physical supplies, eg, potable water, nonperish- (1) ZIP code, which then auto-populates fields for city,
able food, communications equipment, and power gen- state, and county, then have the user continue entry
erators; establishment of relational networks between for (2) street name and (3) house number, and com-
response groups, eg, first responders, local govern- pare typed results against a preloaded database of
ments, and nonprofit aid organizations; and assess- street names and house numbers and either disallow
ment of critical infrastructure, eg, hospitals, electri- nonmatching entries or flag them for review before a
cal grid components, and evacuation transportation record can be completed. Had this been available dur-
routes. While these components remain critical, pre- ing the start of Operation Blue Roof, there would have
paredness now needs to include considerations of data
collection systems, data sets and data standards, and

Special Issue on Puerto Rico

262 Journal of Emergency Management
Vol. 19, No. 8

been efficiency improvements in processing requests nnNumber of on-site assessments completed,
for assistance. relative to the number of requests.

The Federal response and recovery efforts after nnNumber of construction actions or inter-
the 2017 hurricane season exposed gaps in the vention actions, eg, debris removal, com-
completeness of housing address data infrastruc- pleted relative to the number of requests.
ture, especially in Puerto Rico. Subsequent efforts
to improve residential address data collection are nnTime elapsed between request for assis-
ongoing across multiple federal agencies,28 with the tance and assistance provision, including
US Census Bureau providing support to projects such observed minimum time, maximum time,
as the nonprofit Initiative for Civic Address Systems average time, and median time, separable
Assessment in Puerto Rico29,30 to support enhanced by spatial unit. (This feature has a special
local residential address collection. The results of relevance for potential contract, financial
address infrastructure updates should be routinely management, and auditing actions after a
reviewed and incorporated into geospatial databases, mission is completed, which is beyond the
including for rural areas on the US mainland, which scope of this paper but warrants mention
might have historically lacked traditional street because of the relevance to data capture
address. Such data-maintenance protocols should be system design.)
an integral part of any data-collection tool used for
federal response and recovery. nnNumber of intervention actions completed
(or end point reached), relative to the
Principle 2:identify commonly needed metrics ­beforehand number of requests, relative to the popula-
Disasters that are severe enough to involve federal tion, or relative to a defined spatial unit.

assistance will often have recurring types of informa- These metrics are only a starting point; they
tion requirements; this was the case during the extent represent a minimum that should be made readily
of the Operation Blue Roof and TFP missions. Data available to decision-makers from day 1 of any natu-
collection systems should consider these recurring ral disaster response mission. Any team designing
information needs, include nested units of measure, data collection and management systems should be
eg, neighborhood, city, or municipality, county, and made aware of these high-frequency reporting needs,
state, and provide ways to aggregate and disaggre- and the reality that different summaries and visu-
gate data across disaster events. Unique numerical alizations will be required. Common data types and
case identifiers assigned during one event should structures also allow for pre-event development and
not be reused during another event. Each disaster testing of visual summaries with historical or hypo-
response will have three general phases: (1) request thetical data.
for assistance and assessment phase; (2) intervention
action or construction phase; and (3) close-out phase. Principle 3: tools should improvise
Common metrics that should be readily available alongside the m­ ission
and extractable from relevant disaster management
database include: Emergency management incorporates a large
amount of improvization and creativity. Improvization
nnNumber of new requests for assistance does not necessarily indicate the failure of a plan or
received, or processed, per day. the failure to make plans—instead, it is the ability
of the mission to act during the disruptive event.31
nnNumber of requests for assistance relative Miner, Bassoff, and Moorman define improvization as
to the population of the area, or relative to “the deliberate and substantive fusion of the design
the number of homes in an area.

Special Issue on Puerto Rico

Journal of Emergency Management 263
Vol. 19, No. 8

and execution of a novel production,”32 which means time and expertise, and other factors,20,36 and many
that it allows room for conception and implementa- resources exist for comparing and contrasting meth-
tion of action when it is needed.33 Improvization is a ods.37,38 Multiple visualizations are often needed to
desirable quality in any emergency management mis- summarize complex data sets such as those encoun-
sion, and therefore, any emergency data management tered in disaster management, and critical thinking
and visualization system. and trial-and-error may be required pre-emptively
and in real-time to identify appropriate methods
Resilience is the ability to manage uncertain- (Principles 1 and 3). For example, during Operation
ties productively, and to do this, decisions must be Blue Roof, there was a high demand for municipality-
made with the best available information. This flow level data, both to inform local government leaders
of data into information for decision-making is criti- and to identify areas where construction activities
cal to the success of a response and recovery effort. were not proceeding at the expected pace.
Technological advances have resulted in capabilities
to create very powerful and mobile response and A wide array of technical tools can facilitate the
recovery tools. These tools can summarize informa- management, visualization, exploration, and com-
tion vertically within an organization—from the field munication of disaster response and recovery data.
to headquarters—and/or to coordinate across a wide A number of context-specific tools have emerged for
variety of stakeholders, including nonprofits, individ- applications related to humanitarian relief coordina-
uals, and local, state, and federal government agen- tion,39 emergency incident management,39 recovery
cies. However, for such tools to be trusted, information of private businesses,8,40 situation reporting,41 and
must be accurate, up-to-date, and oriented in a form disaster cloud data management.42 Additionally, data
that makes it easy for decision makers to achieve exploration and real-time report generation tools are
situational awareness. Recent research has identified emerging, which could be tailored to ad hoc needs and
that there is a gap between the planned functions of automate key parts of the response process. Notably,
disaster response tools and their performance during many methods are available as Free and Open Source
a disruption because of the failure of supporting infra- Software (FOSS), which has been highlighted as
structure systems like power or communications.34 particularly appropriate for disaster management
Tools need to be developed with similar improvi- applications.43 Two groups of tools are highlighted
zational abilities as the missions they support—­ here that are particularly germane to the role of visu-
specifically reconfigurable and modular technologies.35 alization in informing a disaster response effort: data
By acknowledging potential restrictions to the toolsets dashboards and open science methods.
and the need for improvization, data management
professionals can better design adaptive technologies “Data dashboards” generally describe a graphical
to improve decision-making when it counts the most. user interface summarizing multiple data streams
relevant to decision-makers typically in the form of
Principle 4: data visualization is a series of charts, infographics, and summary sta-
more than a single plot tistics. The public has gained broad awareness of
these portals during the COVID-19 pandemic, during
Information visualization “aims to enhance one’s which many organizations established web services
ability to carry out a task,”5 and thus, the type of for conveying rapidly evolving data on case counts,
visual summary is intimately linked to the disaster mortality, hospital capacity, and local demography
management decision being made.19,20 Furthermore, all across space and time.44 Dashboards provide
visualization connects data to a narrative and an a powerful mechanism for different audiences to
audience’s analytical and emotional responses.21 The explore data through multiple summary figures as
scope of data visualization is typically driven by well as investigating outcomes at multiple scales, eg,
issues of purpose, data type and distribution, con- municipality, county, state, or nation. From a software
straints of the visual media, logistical issues such as developer perspective, these tools provide the ability

Special Issue on Puerto Rico

264 Journal of Emergency Management
Vol. 19, No. 8

for dynamic update as data become available (rather centrally managed electrical generation and distribu-
than manually updated figures pasted into briefing tion network, and the Operation Blue Roof mission
books), capacity to control differential access on public provided temporary roofing to individual homeown-
and sensitive data, eg, through password protection, ers. These two efforts were operating among other
and easy mechanisms to “push” new visualizations or response and recovery efforts managed by other gov-
analytical results out to the audience. ernment, private, and nonprofit entities.

“Open science” refers to an approach embracing Although these missions varied in aim and scope,
transparency in all stages of the research process, the need for accurate data collection, analysis, and
including sharing of data, analytical code or software, communication was constant. During the TFP and
published products, and other issues.45 Disaster man- the Operation Blue Roof mission, data communication
agement often requires sensitive data such as per- evolved in response to increased technical capabilities
sonal information, health data, or financial outcomes, and staff familiarity with data sources and communi-
and thus, a complete embrace of data sharing may be cation needs. Based on the authors’ experience, it is
inappropriate. However, the tools provided by the open believed that if some of the suggested tools and tech-
science community could assist managers in trans- nologies had been in place before the disaster, it could
parently sharing outcomes with interested parties have improved the ability of field staff to perform
ranging from agency leaders and field practitioners data collection, database query processes, and speed
to affected citizens and policy makers. One example of data communication and reporting, but it would not
of an open science application after Hurricane Maria necessarily have sped up the overall response. Time
came from a group of interdisciplinary scientists who is a precious commodity during disaster response,
responded to the health hazard of limited water treat- and it behooves agencies to have staff focusing on
ment capacity due to power outages and subsequent mission delivery and tactical goals, not chasing down
concerns about the risk of leptospirosis. This group typographical errors that could have been prevented.
collected water samples from across the island and For example, switching from the labor-intensive ball-
created an open-source cyber-infrastructure that con- and-stick model of the electrical grid versus the later
nected communities, researchers, and practitioners to maps that were automatically generated (and much
real-time water quality data and tools for analysis.46 less prone to error) through ArcGIS allowed staff to
From a technical standpoint, real-time report gen- focus on other aspects of data collection and reporting.
eration tools like Markdown, Jupyter notebooks, and Figures 5 and 6 display examples of different data vis-
Sweave can facilitate rapid data sharing45 and avoid ualization presentations, and how these could be used
duplication in preparation of situation reports and to understand different types of mission progression
briefings. Furthermore, these methods can be used to (subobjectives). In some cases, data aggregation is use-
share predictive modeling and allow decision-makers ful to display overall progress and discrete numbers to
to interact with models by changing parameters convey the total mission size. Conversely, disaggrega-
related to policy options such as Stanford University’s tion is essential to display individual case details that
COVID-19 modeling toolkit.47 may be critical knowledge for field personnel. During
the post-Hurricane Maria missions, there were subob-
Discussion jectives that required information on the overall trend
in roof installation or power connectivity, ie, the cen-
The need for assistance after natural disasters tral tendency, priority installations for residents with
is unlikely to disappear. These events will vary in special needs, eg, health conditions exacerbated by
the type of damage they inflict but can include net- group shelters, the installation of the “last” roof or last
work impacts (electric, water, sewer, and internet) substation, ie, the extremes of the distribution, and the
and patch-style impacts (individual structures and spatial distribution of installations across the island.
debris). The case studies described above illustrate A single visualization is unlikely to inform all aspects
the range of these issues—TFP mission engaged a

Special Issue on Puerto Rico

Journal of Emergency Management 265
Vol. 19, No. 8

of a complex disaster-response mission; since purpose nnDesign systems for use in environments
should drive the visualization, multiple figures will with limited internet connectivity. Natural
be needed for multipurpose disaster responses, and disasters might interrupt critical commu-
visuals must be interpreted and contextualized to nication elements such as internet cables
effectively inform decisions. and cellular towers. Satellite-based com-
munications are not yet widely available
The scale and type of response missions neces- and should not be expected to serve all
sitates different specialty teams. Personnel deployed potential communication needs.
to the Blue Roof mission did not simultaneously serve
within the TFP mission; data management was also nnReduce the need for manual data entry
separated. We suggest that future data collection, as much as possible. Preloading elements
storage, and management efforts have the capabil- such as address fields are an important
ity to be harmonized, so that multifaceted disaster element of any system since so much
response efforts can receive more complete situational disaster response and recovery is spa-
awareness. In many places, coordinated tracking tially based and may require coordina-
could be done by using a distinctive numerical iden- tion across multiple geographic entities
tifier in addition to a geospatial feature such as an by people who are unfamiliar with local
address, but this would harmonize better across dif- names and spellings.
ferent databases if data system architects agreed on
an address structure in advance. nnInclude database fields for standard and
nonstandard geospatial information.
In this paper, we suggest a list of considera- Some communities may have limited road
tions for those building or managing a data collec- access or low-clearance bridges along cer-
tion system for use during disaster response and tain routes, or hazards like mudslides
recovery operations; these include features related might be present as a result of the disas-
to data input and data outputs that can be used for ter. This type of access-related informa-
management, reporting, and coordination with other tion should be captured with geospatial
stakeholders. In summary, we recommend the follow- information wherever possible.
ing elements to be considered in data management
system design and maintenance:

nnConsult emergency management practi- nnPreload systems with relevant geospatial
tioners during software design and test- datasets. Relevant datasets might include
ing. Practitioners who work in disaster topographical maps, census-block data-
response represent a variety of professions sets, street maps, flood risk maps, and
and have experience with different data other spatially based information such as
sources and presentation styles. Their pro- social vulnerability maps.48,49 Contingency
fessional experience is required for identi- environments may have limited or no
fying what pieces of information are used internet connectivity, so any geospatial
or desired during response management. functions should not rely on internet-
Ongoing consultation between software based connectivity to a remote server.
developers and practitioners is critical These datasets should be reviewed and
because people who may be responsible updated at least annually.
for generating or capturing data, and
downstream data consumers, are often nnDesign reporting systems that will be able
neither information technology specialists to disaggregate and aggregate data in dif-
nor have experience in data management. ferent ways to allow for the evolution of

Special Issue on Puerto Rico

266 Journal of Emergency Management
Vol. 19, No. 8

information needs during a mission. The of plywood or utility poles, but they can communicate
places or questions that are asked at the the urgency in a more visceral way. We hope that this
start of a response mission will evolve as paper will serve as a learning opportunity and guide
more information becomes available and for improved planning and collaboration across the
the recovery process progresses. multiple stakeholders involved in disaster response
and recovery. A key message is that generic data man-
nnInclude database outputs that can be agement systems exist (long-term strategic goals) but
quickly generated and exported into com- are somewhat challenging to tailor in real-time (in
monly used file types, eg, .CSV, .TXT. If the face of real-time, response-driven tactical goals).
search and selection capabilities are not For organizations that regularly respond to disasters,
built into the database itself, sensitive we echo the recommendation that a data manage-
data must be able to be removed before ment specialist position should be embedded in early-
file dissemination. response teams.17 These data management specialist
should be familiar with the data collection system(s)
nnAllow different levels of information as well as the reporting needs commonly associated
access within the database. Different with emergency response, described in the next sec-
response roles will require different lev- tion. This position is critical in supporting the variety
els of database access (permissions). Any of action areas that contribute to a response mission,
dashboards within a data system should from logistical to legal, and might require skillsets,
include relevant contextual data to pro- eg, database management, geospatial analysis, not
mote accurate interpretation of any sum- normally emphasized in early-response teams.
mary v­ isualizations.
Acknowledgments
nnRegularly test software. Computer operat- Prior to publication, this paper was significantly improved
ing systems go through regular updates, from readings by Anne P. Bello, J. Pettit, R. Moser, and two anony-
which may be incompatible with older mous reviewers. Opinions expressed herein are those of the authors
software; this should be examined through and not necessarily those of the agencies they represent. The use of
regular testing of all functions and cor- products or trade names does not represent an endorsement of these
rected as required. products by either the authors or the Department of the Army.

We make these suggestions while cognizant of the Marin M. Kress, PhD, Research Physical Scientist, US Army Engineer
fact that delays in assistance have very real impacts Research and Development Center, Coastal and Hydraulics Laboratory,
to those affected; every Blue Roof case number rep- Vicksburg, Mississippi. ORCID: https://orcid.org/0000-0002-5835-5686.
resented not just a structure, but a home. Every util-
ity pole installed represented another step toward a Katherine F. Chambers, MS, Research Physical Scientist, US Army
community returning to their normal daily rhythms. Engineer Research and Development Center, Coastal and Hydraulics
Although valuable as tools, data visualizations, Laboratory, Vicksburg, Mississippi.
graphs, and charts can have the unintended side
effect of obscuring the human element of a disaster Darixa D. Hernández-Abrams, MS, Research Ecologist, US Army
behind numbers and maps. Aggregated progress bars Engineer Research and Development Center, Environmental
can mislead or omit the nuance of which communities Laboratory, Vicksburg, Mississippi.
are still struggling to recover and why. Objectivity
in disaster management cannot replace the original S. Kyle McKay, PhD, PE, Research Civil Engineer, US Army Engineer
goal of providing humanitarian assistance. The best Research and Development Center, Environmental Laboratory, New
graphics in the world do not change the availability York, New York. ORCID: https://orcid.org/0000-0003-2703-3841.

REFERENCES
1. Gunderson L: Ecological and human community resilience in
response to natural disasters. Ecol Soc. 2010; 15(2): 18.
2. IPCC: Climate change 2014: Synthesis report. Contribution of
working groups III and III to the fifth assessment report of the
intergovernmental panel on climate change. 2014.

Special Issue on Puerto Rico

Journal of Emergency Management 267
Vol. 19, No. 8

3. Pelling M: The Vulnerability of Cities: Natural Disasters and www.usace.army.mil/Missions/Emergency-Operations/National-
Social Resilience. London: Earthscan Publications Ltd., 2003. Response-Framework/Temporary-Roofing/. Accessed September 1, 2020.
4. Brown T, Craddock S, Ingram A: Critical interventions in global 24. Rosati JD, Touzinsky KF, Lillycrop WJ: Quantifying coastal
health: Governmentality, risk, and assemblage. Ann Assoc Am Geogr. system resilience for the US army corps of engineers. Environ Syst
2012; 102(5): 1182-1189. DOI: 10.1080/00045608.2012.659960. Decis. 2015; 35(2): 196-208. DOI: 10.1007/s10669-015-9548-3.
5. McInerny GJ, Chen M, Freeman R, et al.: Information visualisa- 25. Sharifi A, Yamagata Y: A conceptual framework for assessment
tion for science and policy: Engaging users and avoiding bias. Trends of urban energy resilience. Energy Procedia. 2015; 75: 2904-2909.
Ecol Evol. 2014; 29(3): 148-157. DOI: 10.1016/j.tree.2014.01.003. DOI: 10.1016/j.egypro.2015.07.586.
6. Yu M, Yang C, Li Y: Big data in natural disaster management: A 26. Heinimann HR: A generic framework for resilience assessment.
review. Geosciences. 2018; 8(5): 165. DOI: 10.3390/geosciences8050165. In Resilience Resource Guide: An Edited Collection of Authored
7. Naumann F, Raschid L: Information integration and disaster Pieces Comparing, Contrasting, and Integrating Risk and Resilience
data management (DISDM). 2006. Available at http://db.cis.upenn. with an Emphasis on Ways to Measure Resilience. Geneva: EPFL
edu/iiworkshop/postworkshop/positionPapers/123.pdf. Accessed International Risk Governance Center, 2016: 90-95. DOI: 10.5075/
January 11, 2020. epfl-irgc-228206.
8. Hristidis V, Chen S-C, Li T, et al.: Survey of data management 27. Wardekker A, Wilk B, Brown V, et al.: A diagnostic tool for sup-
and analysis in disaster situations. J Syst Softw. 2010; 83(10): porting policymaking on urban resilience. Cities. 2020; 101: 102691.
1701-1714. DOI: 10.1016/j.jss.2010.04.065. DOI: 10.1016/j.cities.2020.102691.
9. Pasch RJ, Penny AB, Berg R: National Hurricane Center tropical 28. Puerto Rico Address Data Working Group (PRADWG) of
cyclone report: Hurricane Maria. Tropical cyclone report AL152017. the Federal Geographic Data Committee Address Subcommittee:
National Hurricane Center, 2018. Phase 2 report: Federal agency challenges in the management of
10. Cangialosi JP, Latto AS, Berg R: National Hurricane Center tropi- Puerto Rico Address Data. Report to the Federal Geographic Data
cal cyclone report: Hurricane Irma. National Hurricane Center. 2018. Committee. 2020. Available at https://communities.geoplatform.
11. Kishore N, Marqués D, Mahmud A, et al.: Mortality in Puerto gov/ngda-address/wp-content/uploads/2020/05/PRADWG-Phase-2-
Rico after Hurricane Maria. N Engl J Med. 2018; 379(2): 162-170. Report-FINAL.pdf.
DOI: 10.1056/NEJMsa1803972. 29. iCASAPR: About iCASAPR. 2019. Available at https://icasapr.
12. Rivera R, Rolke W: Modeling excess deaths after a natural dis- com/. Accessed September 1, 2020.
aster with application to Hurricane Maria. Stat Med. 2019; 38(23): 30. US Census Bureau: The opportunity project: Puerto Rico
4545-4554. DOI: 10.1002/sim.8314. address database operations support (PRADOS). 2020. Available at
13. National Hurricane Center: Costliest US tropical cyclones https://opportunity.census.gov/prados/. Accessed September 1, 2020.
table updated. 2018. Available at https://www.nhc.noaa.gov/news/ 31. Kreps GA, Bosworth SL: Organizational adaptation to disaster.
UpdatedCostliest.pdf. Accessed January 11, 2020. In Handbook of Disaster Research. Berlin: Springer, 2007: 297-315.
14. US Army Corps of Engineers: Emergency Response Fact Sheet. DOI: 10.1007/978-0-387-32353-4_17.
2011, 1. Available at www.usace.army.mil. Accessed September 1, 2020. 32. Miner AS, Bassoff P, Moorman C: Organizational improvisation
15. US Government Accountability Office: 2017 Hurricane season: and learning: A field study. Adm Sci Q. 2001; 46(2): 304-337. DOI:
Federal support for electricity grid restoration in the US Virgin 10.2307/2667089.
Islands and Puerto Rico (Gao-19-296), 2019. 33. Kendra J, Wachtendorf T: Improvisation, creativity, and the art
16. Li T, Xie N, Zeng C, et al.: Data-driven techniques in disaster of emergency management. In Durmaz H, Sevinc B, Yayla AS, et al.
information management. ACM Comput Surv. 2017; 50(1): 1-45. (eds.): Understanding and Responding to Terrorism. Amsterdam:
DOI: 10.1145/3017678. IOS Press, 2007: 324-335. Available at https://books.google.com/book
17. National Research Council: Improving Disaster Management: s?id=jZjqavWoRfkC&lpg=PA324&ots=GzBgEUUcOq&dq=(Kendra
The Role of IT in Mitigation, Preparedness, Response, and Recovery. and Wachtendorf 2007&lr&pg=PA324#v=onepage&q=(Kendra and
Washington, DC: National Academies Press, 2007. Wachtendorf 2007&f=false. Accessed September 1, 2019.
18. Raschid L: Information integration and dissemination for 34. Tim Y, Pan SL, Ractham P, et al.: Digitally enabled disaster
disaster data management. In Proceedings of the 8th Annual response: The emergence of social media as boundary objects in
International Conference on Digital Government Research: Bridging a flooding disaster. Info Systems J. 2017; 27(2): 197-232. DOI:
Disciplines & Domains. Los Angeles, CA: Digital Government 10.1111/isj.12114.
Society of North America, 2007: 333-335. Available at https://dl.acm. 35. Singh JB, Sharma SK, Misra S: “Invented-on-the Fly” mobile
org/doi/abs/10.5555/1248460.1248543. Accessed May 1, 2020. application for disaster response: Construction of technologi-
19. Kelleher C, Wagener T: Ten guidelines for effective data visuali- cal frames and impact. EGOV-CeDEM-ePart 2019. 2019: 99-108.
zation in scientific publications. Environ Model Softw. 2011; 26(6): Available at https://biblio.ugent.be/publication/8626904/file/
822-827. DOI: 10.1016/j.envsoft.2010.12.006. 8626906.pdf#page=113. Accessed January 11, 2022.
20. Wang D, Guo D, Zhang H: Spatial temporal data visu- 36. McKay SK: Visualization as a tool for ecological analysis. In
alization in emergency management. In Proceedings of the Encyclopedia of Ecology. Amsterdam: Elsevier, 2019: 213-220. DOI:
3rd ACM SIGSPATIAL Workshop on Emergency Management 10.1016/B978-0-12-409548-9.10566-4.
Using-EM-GIS’17. New York: ACM Press, 2017: 1-7. DOI: 37. Tominski C, Aigner W: The TimeViz browser: A visual survey of
10.1145/3152465.3152473. visualization techniques for time-oriented data. 2015. Available at
21. Engebretsen M, Kennedy H (eds.): Data Visualization in http://survey.timeviz.net/. Accessed September 1, 2020.
Society. Amsterdam: Amsterdam University Press, 2020. DOI: 38. Lengler R, Eppler MJ: A periodic table of visualization methods.
10.5117/978946372290221. 2020. Available at www.visual-literacy.org/periodic_table/periodic_
22. Lloyd J: Department of Energy Project Management Workshop: table.htm. Accessed September 1, 2020.
Puerto Rico grid restoration presentation slides. 2018. 39. Sahana Software Foundation: Sahana foundation: Open source
23. US Army Corps of Engineers: Emergency operations: National disaster management software. 2020. Available at https://sahana
Response Framework: Temporary roofing. 2020. Available at https:// foundation.org. Accessed September 1, 2020.

Special Issue on Puerto Rico

268 Journal of Emergency Management
Vol. 19, No. 8

40. Florida International University: Business continuity infor- 44. Johns Hopkins University & Medicine: Coronavirus resource
mation network. 2020. Available at http://www.bizrecovery.org/. center. 2020. Available at https://coronavirus.jhu.edu/us-map.
Accessed September 1, 2020. Accessed September 1, 2020.
41. Zheng L, Shen C, Tang L, et al.: Disaster SitRep—A vertical 45. Hampton SE, Anderson SS, Bagby SC, et al.: The tao of open
search engine and information analysis tool in disaster manage- science for ecology. Ecosphere. 2015; 6(7): art120. DOI: 10.1890/
ment domain. In 2012 IEEE 13th International Conference on ES14-00402.1.
Information Reuse & Integration (IRI). Piscataway, NJ: IEEE, 2012: 46. Hart J, Bandaragoda C, Ramirez-Toro G: Sharing data helps
457-465. DOI: 10.1109/IRI.2012.6303044. Puerto Ricans rebound after Hurricane Maria. Eos (Washington
42. Grolinger K, Capretz MAM, Mezghani E, et al.: Knowledge DC). 2019; 100. DOI: 10.1029/2019EO121651.
as a service framework for disaster data management. In 47. Stanford University: Systems utilization research for Stanford
2013 Workshops on Enabling Technologies: Infrastructure for medicine: COVID-19 tools. 2020. Available at https://surf.stanford.
Collaborative Enterprises. Piscataway, NJ: IEEE, 2013: 313-318. edu/covid-19-tools/covid-19/. Accessed September 1, 2020.
DOI: 10.1109/WETICE.2013.48. 48. Cutter SL: Vulnerability to environmental hazards. Prog Hum
43. Careem M, De Silva C, De Silva R, et al.: Sahana: Overview of a Geogr. 1996; 20(4): 529-539. DOI: 10.1177/030913259602000407.
disaster management system. In 2006 International Conference on 49. Cutter SL, Barnes L, Berry M, et al.:A place-based model for under-
Information and Automation. Piscataway, NJ: IEEE, 2006: 361-366. standing community resilience to natural disasters. Glob Environ
DOI: 10.1109/ICINFA.2006.374152. Chang. 2008; 18(4): 598-606. DOI: 10.1016/j.gloenvcha.2008.07.013.

Special Issue on Puerto Rico

Journal of Emergency Management 269
Vol. 19, No. 8

Notes

Special Issue on Puerto Rico
270 Journal of Emergency Management

Vol. 19, No. 8

Bring America’s Foremost Leading the Way in Disaster Medicine
Disaster Medicine Experts
To Your Next Disaster Call. e American Journal of Disaster Medicine is edited,
written and peer-reviewed by the most impressive
Vol. 3, No. 3–May/June 2008 and knowledgeable names in the eld, including the
ISSN 1932-149X leading innovator in disaster medicine,
Susan Briggs, M.D., our editor-in-chief. So you know
AMERICAN JOURNAL OF you’re getting the most accurate, up-to-date and
relevant information in every issue. Indexed in PubMed
DISASTER and Medline, no other publication o ers this level of
MEDICINETM expertise in this unique specialty.

Official Journal of the American Society of Disaster Medicine Specializing in Your Specialty

CONTENTS Published quarterly, every issue of the
American Journal of Disaster Medicine provides
Original Articles • The performance-level model in-depth coverage on injuries and illnesses
• Disaster medicine: A history ................125 of hospital decontamination most physicians have never seen—like blast wounds
preparedness ......................................157 and post-explosion injuries, biological weapons
Geoff Williams contamination and mass physical
Scot Phelps, JD, MPH and psychological trauma.
• The impact of natural Garrett Doering, MS
disasters on the elderly ........................133 Getting the Edge in Saving Lives
• The impact of correctional
E. Paul Cherniack, MD institutions on public health during Put America’s leading experts in disaster
a pandemic or emerging medicine on your disaster medicine team.
• A unique role for dental school infection disaster ..................................165 Subscribe to the American Journal of Disaster
faculty: Telephone triage training Medicine by calling 1-800-743-7206, visiting
and integration into a health Rachel D. Schwartz, PhD www.disastermedicinejournal.com or lling out
departments’ emergency the enclosed form. Subscribe today!
response planning ..............................141 • Hospital-based disaster
preparedness for pediatric Learn from the Leading Innovators
Jill B. Fernandez, RDH, MPH patients: How to design a realistic in Disaster Medicine, with the
David L. Glotzer, DDS set of drill victims ....................................171 American Journal of Disaster Medicine
Marc M. Triola, MD
Walter J. Psoter, DDS, PhD Shana Ballow, DO 4 WAYS TO SUBSCRIBE:
Solomon Behar, MD
• The impact of the Tsunami on Ilene Claudius, MD WEB www.disastermedicinejournal.com
hospitalizations at the tertiary care Kathleen Stevenson, RN PHONE 800-743-7206 x108 (Toll Free)
hospital in the Southern Province Robert Neches, PhD
of Sri Lanka ........................................147 Jeffrey S. Upperman, MD, FAAP, FACS 781-899-2702 x108 (Outside USA)

Truls Østbye, MD, PhD • International Disaster Response MAIL 470 Boston Post Road, Weston, MA 02493
Thyagi Ponnamperuma, MBBS Law: An introduction ..............................181
Nayana Fernando, MBBS, MSc FAX 781-899-4900 (24 hours/day)
Vathsala Abeygunawardena, MSc Lewis W. Marshall, Jr, MS, MD, JD
W. A. A. Wijayasiri, MBBS, MSc
Jodi J. Chen, MD
Christopher W. Woods, MD, MPH

Indexed in PUBMED/NLM/DMIRC/SCOPUS

JOURNAL OF EMERGENCY MANAGEMENT

Preparing You for
Whatever Disaster Comes Your Way

The Most Pressing Issues The Leader in Emergency Management
in Our Industry
The Journal of Emergency Management is written, edited and
Every issue of Journal of Emergency Management is peer-reviewed by an internationally recognized team of the
peer-reviewed and packed with invaluable information foremost hands-on EM experts. They include top professionals
and insight. Published bi-monthly, topics include: from the public and private sectors who offer real world
„ Emergency planning and response experience and practical solutions, as well as leading academics
„ Disaster recovery and business continuity planning who provide perspective and analysis on the latest research and
„ Emergency preparedness and response legislation studies. Together, they bring you the most thorough, relevant
„ Risk management and useful source of information on emergency management.
„ Emergency management today, tomorrow and the future And, Journal of Emergency Management has been accepted
„ 6HYHUH ZHDWKHU ÁRRG DQG KXUULFDQH VWXGLHV for indexing in the National Library of Medicine’s prestigious
„ (PHUJHQF\ QRWLÀFDWLRQ DQG FRPPXQLFDWLRQ PUBMED database as well as the DMIRC and SCOPUS.
„ Continuity of operations and infrastructure protection
„ Preparation and evacuation of special populations The Most Important Thing You Can
„ Cross-training in emergency management Do for Yourself and Your EM Team
„ And much more, including the latest on COVID-19
Make sure you and your team stay one step ahead of disaster.
Available in print, online and site license access Subscribe today by calling 1-800-743-7206
or visit www.emergencymanagementjournal.com.
ISSN 1543-5865
Subscribe now, and be prepared for anything!

4 Ways to Subscribe

WEB www.emergencymanagementjournal.com

PHONE 800-743-7206 x108 (Toll Free)
781-899-2702 x108 (Outside USA)

Mail 470 Boston Post Road, Weston, MA 02493

Fax 781-899-4900 (24 hours/day)

www.emergencymanagementjournal.com