County of Hawai‘i Multi-Hazard Mitigation Plan - Our Comment

WITH PHILIP ONG SUMMERY This review and comment was submitted to Tetra Tech to help improve the Hawai‘i Multi-Hazard Mitigation Plan before being fully adopted by the County of Hawaii. The plan itself is very robust, and represents a significant step forward for disaster mitigation on our island. Some of the issues identified in the document relate to; the 2018 Summit Collapse Events on Kīlauea, needed or incomplete references, wildfire zoning areas, and proper consideration of cyber, a few hazards relating to our island’s volcano, and a few more smaller details. Tetra Tech and Hawaii County Civil Defense did take the time to review these comments with Philip Ong and myself, we appreciate that effort to hear us out. ----- Review Introduction Tetra Tech has been working with the County of Hawai‘i to develop a Multi-Hazard Mitigation Plan (HMP) that examines a variety of disasters in the County of Hawai‘i, or more simply, Hawai‘i Island. To start, the 2020 draft plan created by Tetra Tech is a great improvement on the previous plan adopted by the County in 2015, and it includes sound recommendations, conclusions, and an action plan for the County to follow for obtaining funding and emergency aid. We are providing a formal pro-bono review of the draft 2020 HMP. For the sake of transparency, we now publicly report the findings that arose in our independent analysis. None of the issues we identify compromise the final analysis or the proposed action plan, but should be corrected or addressed. Due to the short review time and the document’s length, we do not address every small aspect of the HMP. Our intent is to improve the document to be more thorough and complete, as future HMPs can better address hazards that are acknowledged in the current version. Feedback on Methods The HMP methodology focuses on hazards for the next 100 years, but naturally relies most heavily on records from the past 100 years. This works well with most data that have inherently shorter time cycles and higher frequencies, such as tropical cyclones and flooding which can be impactful on a yearly basis. The issue is that the most recent 100-year window is only a single data sample from which to draw the bulk of the conclusions, but there exist other 100-year window data samples (such as from 1820-1920) which have limited, but useful information which could improve the analysis. The approach taken catches most hazards, but is not 100% complete. This is especially true for the volcanic risks which cycle on longer time-scales, where impactful hazards which may only occur once every 500 years but warrant analysis and inclusion in a comprehensive document. While older events are mentioned in the HMP — the eruption of Mauna Loa in 1881 that fed lava into Hilo, the 1859 radial eruption of Mauna Loa that reached the sea, the tuff cone at Kapoho Crater, and eruptions on Hualālai that threaten North Kona — they are not fully considered as potential future threats beyond inclusion in the analysis of historical lava flows. A larger emphasis on the accompanying Volcanic Risk Assessment would ease such concerns. From an average citizen’s perspective, another concern is that this HMP appears to put more focus on mitigating damages to property and the carry-on effect of property damage on people, and not on mitigating the direct effect on people themselves during a disaster, especially one that lasts for several months. This could be easily addressed by adding emphasis that the HMP will be used to support the Emergency Operations Plan that focuses on the immediate health and safety of the community during a disaster. In general the HMP should be addressed to the public and not only to FEMA and define its “mitigation” terminology and scope early (perhaps in the executive summary, the section a member of the public is most likely to read). The HMP states one of its primary purposes is to be used as guide and reference to seek sources of funding through grants and other sources of revenue to support the needs that will arise in various disasters. It is successful in laying out an easy to understand plan for the County to follow that also satisfies its FEMA requirements. However, applying this type of “template” to our unique island can sometimes produce bizarre results, or more concerning omissions, as detailed below. Missing Content in the HMP - Clarification on Natural and Non-Natural Hazards RECOMMENDATION: Add additional clarification on the non-natural classification. Include text stating benefit of performing more detailed assessments of non-standard hazards in future HMPs. ISSUE: In section 6.1, the draft text lists Invasive Species & Pandemic Outbreaks as “non-natural hazards”, apparently due to limitations on full assessments of those hazards. The source of wildfires is often man-made, yet it is listed as a natural hazard, whereas a pandemic or invasive species are not. Even without further introductions by humans, contagions and invasive species that are already present on the island represent a threat to local ecosystems, food production, population health, and the economy and deserve greater future assessments. While the Invasive Species section has now been deemed “natural” via public presentation, it is not fully assessed as a consequence of limited available data. Further, the listing of a pandemic as a non-natural hazard creates a potential for unwarranted speculation in public interpretation due to the emergence of SARS-CoV-2 and theories online regarding the Wuhan Institute Virology. To account for the possibility of “natural hazards with limitations on assessment” (invasive species, pandemic) combined with the existing true “non-natural hazards” (supply chain, mass events & cyber) consider a different term that is both acceptable by FEMA and understandable to the general public, perhaps something along the lines of “non-standard hazard”? 2018 Summit Collapse Events As Seismicity & A Separate Sub-Hazard - RECOMMENDATIONS: 10.2.1 Append table 10-4 caption and corresponding text to either specify “collapse-associated events are excluded” or to include them. Update seismic map to show all collapse events with energy of a >M5 earthquake, as well as all offshore events (included). Add subsection “2018 Kīlauea Caldera Collapse” below “2006 Kiholo Bay” based on USGS text included below. 10.2.3 Address collapses as infrequent events consisting of nearly constant earthquakes for a duration of several months. Update or exclude 2015 prediction of >M6.5 in the next 10 years as it already happened. "The modern USGS catalog lists 46 earthquakes on or near the island with magnitudes of 5.0 or greater since 1959, as listed in Table 10-4 and shown on Figure 10-3. Table 10-5 lists historical (pre-1959) earthquakes of magnitude 6.0 or greater. The following sections describe significant earthquakes in the Island’s history." (HMP, pg 10-6) USGS-HVO text: “A total of 62 collapse events occurred between May and early August. Seismic, infrasonic, and geodetic signals settled into a notably consistent pattern characterized by Mw 5.2 to 5.4 VLP collapse events occurring almost daily, with intervening escalating earthquake swarms exceeding 700 earthquakes of magnitude ≤ 4.0 per day. During this phase, higher-frequency seismic energy and strong infrasound signals accompanied the notably consistent VLP signal originally associated with explosions, and the caldera floor dropped several meters during each of the large events, ultimately deepening in places by more than 500 m. The persistent high levels of seismicity caused substantial damage to infrastructure in Hawai‘i Volcanoes National Park, including to the by-then-evacuated HVO facility.” (Neal et al, Science, 2019) "The modern USGS catalog lists 46 earthquakes on or near the island with magnitudes of 5.0 or greater since 1959, as listed in Table 10-4 and shown on Figure 10-3. Table 10-5 lists historical (pre-1959) earthquakes of magnitude 6.0 or greater. The following sections describe significant earthquakes in the Island’s history." (HMP, pg 10-6) ISSUE: Not all 2018 events are accounted for. Of the 62 collapse events, 54 are documented in the USGS catalog under the “Volcanic Eruption” classification with magnitudes from 5.0 to 5.4 that should be included as part of the Ch.10 Earthquake analysis. They were felt widely outside the National Park and caused persistent localized damage to critical roadways and nearby neighborhoods, specifically Volcano Village, who remain exposed to future events. Furthermore, it’s worth considering that while previous HMPs did not anticipate or include this “earthquake style” as a hazard, that knowledge has been gained from the events in 2018 and should be exported. As it stands, the possibility of further summit collapse on Kīlauea is acknowledged but not extensively discussed in the current plan; these proposed changes address the main effects. Added note: another 6 earthquakes visually appear to be missing from the analysis based on USGS search results and Figure 10-3. Limited consideration of the South Flanks of Kilauea and the ERZ - RECOMMENDATION: Add text noting that further information can be found in the Volcanic Risk Assessment currently pending release. ISSUE: There are only two mentions of the South Flank of Kilauea, even though it is often cited as one of the primary drivers for activity on Kīlauea's East Rift Zone. The May 4th M6.9 at the onset of the 2018 eruption changed the dynamics of the eruption and volcano for years, as over two years later there is continued elevated seismicity along the South Flank at nearly twice the pre-2018 rate. The South Flank deserves the proper consideration historical records and USGS assessments. Coastal Subsidence events in the LERZ - RECOMMENDATION: If further addressed in the Volcanic Risk Assessment then add a sentence to reference the document. Otherwise, denote that further hazard analysis maybe warranted in the future additions to the HMP. ISSUE: Subsidence was directly addressed only in relation and connection to volcanic eruption in Ch.17 (p.242, 247), and is listed as considered on p.416 of the working group but was not included in the HMP. On-island, the south shore of Kīlauea is particularly vulnerable to subsidence as its south flank continues to move seaward, causing widespread subsidence from Kapoho to Halape. Coastal subsidence is mentioned in association with discrete earthquakes but not given further treatment — despite subsidence connected to earthquakes in 1868, 1924, and 1975, significant subsidence has occurred from flank movement in between these discrete events. The County has had to realign several segments of Highway 137 due to the combination of ground subsidence and coastal erosion, especially in the area of Pohoiki Boat Ramp in Puna. Proper Consideration of Lava Zones 3 & 4 (Hualālai, Mauna Loa) - RECOMMENDATION: Add text to 17.3 noting that further information on Lava Zones 3 & 4 can be found in the Volcanic Risk Assessment currently pending release. ISSUE: There are geologic trends that do not adhere to such arbitrary time frames such as centuries especially when making proper considerations on Hualālai, in Lava Zone 4. Hualālai last erupted in 1801, with no eruption in the time of modern instrumentation. While the Tetra Tech HMP does accurately represent two distinct watersheds on Hualālai, and several times includes that the threat of an eruption is high, those considerations are not then carried into the analysis in the highly likely event of a future eruption. The steep topography around the USGS defined summit and rift zones poses significant risk to the areas of Kailua-Kona. The amount of unknowns surrounding an eruption from Hualālai is significant, as noted by the high-end estimates noted in the HMP. Hualālai represents a wild card in any analysis; the chance of massive impact is likely more of a coin flip, when Hualālai does erupt, compared to the more frequent but less impactful on-average eruptions on Mauna Loa, which divides the odds more similarly to a roulette wheel. For future HMPs, USGS-HVO should be consulted further about the high potential for an eruption on Hualālai and the different scenarios that would be possible, specifically concerning how that might impact North Kona. The upper slopes of Hilo exposed to Mauna Loa, within Lava Zone 3, deserves similar consideration. No Consideration of Tuff Cones - RECOMMENDATION: If further addressed in the Volcanic Risk Assessment then no action is needed. Otherwise, denote that further hazard analysis maybe warranted in the future additions to the HMP. ISSUE: Kapoho Crater dates back roughly 500 years, and presents a unique feature on the Lower East Rift Zone of Kīlauea. As opposed to the more common effusive cinder cones, the explosive formation of an ash and tuff cone involving shallow groundwater or sea-water creates unique hazards for the LERZ and communities nearby. The consideration for explosive activity seems to be limited to summit regions of the volcanoes, whereas these interactions can also occur where each active volcano’s rift zones intersect the coast. Wild Fire Zoning and Areas Covered in Lava - RECOMMENDATION: Make note that the only wildfire GIS layer available is outdated and worthy of revisiting by the County of Hawai‘i. If possible, add recommendations on how to secure grants for such a project. Note subdivisions shown on the map have been changed by lava flows since the GIS layer was made; Royal Gardens, Kapoho, Vacationland, Leilani Estates, Lanipuna Gardens, and Kalapana. ISSUE: The GIS source used to derive “Communities at Risk of Wildfire” includes areas that were covered with lava from Kīlauea since the cited reference layer was completed. Royal Gardens, Kapoho, etc, are all no longer a fire risk - as there is no vegetation to burn after being inundated by lava. Water Catchment in Drought- RECOMMENDATION: Add text to 9.4.1 citing County water spigots as a resource for homeowners. Add text to 9.4.2 citing residential catchment tanks are a fire-fighting resource, and drought may affect Fire Department response in rural areas. "Economic impact will be largely associated with industries that use water or depend on water for their business. For example, landscaping businesses were affected in the droughts of the past, as the demand for service significantly declined because landscaping was not watered. Agricultural industries will be impacted if water usage is restricted for irrigation" (HMP, 9.4.1) ISSUE: In section 9.4.1, no mention is made that a large percentage of the County’s rural population use water catchment systems for all of their home water, and as a fire-fighting resource for communities without access to County water. As County water spigots are available for residents during drought, the residential impact is much less than that on the fire-fighting capacity. While this information is present in the mitigation Best Practices, this is unclear beforehand and can be mentioned here, with deferral of management to the Fire Department. Proper Consideration of Cyber - RECOMMENDATION: Indicate the current status a County of Hawaii Cyber Security Action Plan. If no cyber action plan exists for the county, what potential sources of funding are available. ISSUE: The section on cyber threats does create an adequate threat landscape for our interconnected world, though the specifics of online threats are not clearly defined. There are no recommendations or exploration into mitigation measures, best practices, or incident response overviews. The only recommendation made specific to an attack vector is in regards to ransomware (software that ransoms the data of an organization) is to pay the cyber criminals the requested monetary amount to buy back information. Cyber security in general is undervalued until it is too late, only once an organization has been breached does it realize the value of proper security protocols and practices. The concepts of Defense-in-Depth and Layered Security should considered in the report as general recommendations to the County for mitigating cyber threats. The cyber section also does not address new found increases in cyber attacks with COVID-19 as a new consideration. Google has reported that they are detecting roughly 18 million phishing and malware attacks related to COVID-19 each day. Bad actors online are intimidating official organizations in mass to scam and manipulate a fearful population. A full analysis of Cyber as provided in the Section 22 Tables (all non-natural events are not addressed by the scope), MITIGATION BEST PRACTICES, should eventually be amended in the future with a more complete Defense-In-Depth approach to network security for County of Hawai‘i infrastructure. The County’s financial and operational vulnerability is likely significant in comparison to the fully assessed Cyber non-natural hazards in the HMP. Carrington Level Event - RECOMMENDATION: Add text at end of 19.4 intro that protecting against cyber threats has the added benefit of offering some level of protection against natural electro-magnetic threats. If possible, explore this further in future HMPs. ISSUE: Geomagnetic storms could cause significant and extended disruptions to electrical grids and global communications. The last such significant event was the 1859 Carrington Event that caused telegraph stations across the US and Europe to fail, some of which burst into flames, and were able to send messages while the power was disconnected. Recent analysis estimates the odds of another Carrington level event at roughly 0.7% a year. (Chapman, Horne, Watkins, 2020) For Context, From 11.2.7 of the HMP: “Data compiled over the last 50 years indicate that, on average, a damaging flood event occurs on the Big Island with an annual probability of 0.5 percent” - The annual chance of a CME impacting earth is potentially greater at ~0.7%. There have also been near misses and smaller scale geomagnetic events in the last 100 years worthy of acknowledgment in a fully comprehensive risk assessment, even if not considered initially by the HMP working group. The HMP could subtly acknowledge the potential hazards from a Coronal Mass Ejection (CME) for future consideration in revisions to the hazard plan. The analysis could include contingency planning for a limited on-island power supply paired with disrupted connections and global communications brought by a large scale CME. Electromagnetic Pulses (EMP) are mentioned in the section on Cyberterrorism (19.4.2 of the HMP), but natural sources for similar events, such as a CME from the sun, are not considered in the analysis. While EMPs have a higher-frequency pulse that will likely impact impact a wider range of electronic devices and infrastructure, a large scale low-frequency CME could potentially impact anything with an antenna, including the power grid. (Kappenman, 2012) Improper Citations, Poor References and Errors - (Oregon State University no date) "Climate change also could affect recovery of the environment after a volcanic event. For example, vegetation destroyed by volcanic activity takes time to recover, and the length of recovery is dependent on the amount of rain and changes in the climate that the area is experiencing (Oregon State University no date)." REFERENCE TABLE: Oregon State University (OSU). No date. “Volcano World.” A webpage of the Oregon State University website. Last accessed May 2015. Accessed at: http://volcano.oregonstate.edu/ Improper reference and citation. Link does not go to the source material being cited and “Volcano World” is a running publication from OSU. Statement is factually sound. ---- (Danielson, 2015) Table 19-2 The reference used does not show in the reference table. Upon further investigation of the source, it is a writer for Yahoo Finance that was covering a statement made by Joseph Bonavolonta, the Assistant Special Agent in Charge of the FBI. ---- (Oskin, 2012) A tsunami in Hawai‘i can be generated by a nearby as well as a distant earthquake or landslide. Several scenarios could create large tsunami events and greatly impact Hawai‘i County. One scenario includes a local tsunami event triggered by a collapse of a flank of Mauna Loa or Kīlauea in Hawai‘i County. Review of submarine geology indicates historical landslides along these flanks have occurred, and the possibility of another flank collapse is possible. This would probably be very damaging, giving little or no warning time. This could result in great loss of life and property and cause severe environmental impacts (Oskin, 2012). Source is not included in the reference section. ---- (NOVA, n.d.) The USGS categorizes volcanoes as active, dormant, or extinct: • Active volcanoes are generally those currently erupting or showing signs of unrest, such as gas emission and unusual earthquake activity. • Dormant volcanoes are not currently active, but could become restless or erupt again. • Extinct volcanoes are those scientists consider unlikely to erupt again. Previously active volcanoes that formed the older islands of the Hawaiian archipelago are now considered dormant (NOVA, n.d.). This is a poor reference. There is no reference for USGS’s categories for volcanoes, instead the reference provided is to a generic Hawaii volcano NOVA video specified for grades 6-12. The citation is made without a date, and makes a blanket statement about the State of Hawai‘i and dormant volcanoes - which is outside the scope of the HMP. ---- (Swanson and USGS, 2019) REFERENCE TABLE: "Swanson, Don and USGS. 2019. Gather information and data related to volcanic hazard and risk on Hawai‘i (Interview with Tetra Tech)." Don Swanson is the best possible resource to have consulted on the referenced topics. However, there are six references to a conversation that has not been documented or provided. Independent interpretation of the guidance made by Don Swanson cited in the HMP cannot be made, and support for claims made in the text assessed. ---- Tsunamis Table 16-1 Omits the 1946 event referred to throughout the text, the most fatal. ---- Appendix E Tables Several of the tables are confusing in showing too little distinction in impacts by district to lend confidence, yet more detailed information has been presented in map format. For example, there is no difference in the impact on people and property between Puna and North Kohala from the M7.7 Kalapana earthquake, as previously shown by the map on p.142. As this suggests a greater earthquake impact, it does err on the side of caution, but merits some explanation. Perhaps the threshold for seismic impact is so low that the whole island is equally at maximum exposure? ---- Conclusion - The 2020 draft HMP is a great step forward from the County of Hawai‘i’s 2015 HMP, and it can be even more comprehensive with minor additions in certain key areas. The County is a dynamic island with a unique balance of concerns, so it is only appropriate that its HMP should also uniquely address them in a forward-thinking manner, evolving from its experience with recent disasters. The standard process has produced a nearly-complete plan and is designed to be improved by local knowledge and input, which was this review’s goal to provide. The proposed action plan in the short term is unaffected by these additional requested inclusions, whose longer-term planning could influence the maintenance phase of the 2020 HMP and the development of future HMPs.