Page 8 - gt1602

600 4Advance rate (m/day)inflation Summit tilt (microradians) never interrupted), and the Pāhoa solid waste transfer station
500 3 deflation
GSA TODAY | FEBRUARY 2016 2 many residents—perhaps 10%–15% of the area’s population—
400 1
300 0
200 -1
-2
100 -3
0 -4
Sep Oct Nov Dec Jan
Figure 5.Advance rate of June 27th lava flow (blue) versus tilt at Kīlauea’s
summit measured radial to the caldera (azimuth 327°; red). Positive tilt indi­
cates inflation and negative, deflation.Tiltmeter location given in Figure 1.
mayor, HCCD, and, occasionally, local and national elected
representatives were followed by individual interactions among
members of the public, HVO staff, state and county agencies,
and many other groups (the Red Cross, the Federal Emergency
Management Agency [FEMA], Hawai‘i Volcanoes National
Park, etc.).
The meetings built trust among scientists, managers, and the
community and helped residents and businesses understand the
products used to show the flow location and likely pathways.
Equally important, the meetings gave residents a perspective
on the nature of pāhoehoe flows and limitations in forecasting
flow activity.The most important products were current lava
flow maps that included potential flow paths (Fig. 4) based
on topographic steepest-descent calculations (Kauahikaua et
al., 2003; Kauahikaua, 2007). Online release of the maps was
eagerly anticipated by the community, and HVO and HCCD
devoted significant resources to collecting lava flow data,
developing products for rapid public release, and updating the
products in response to user feedback.
Estimation of the lava’s likely time of arrival at a given area
was the most unreliable forecasting parameter because the
advance rate varied rapidly between 0 and 500 m/day (Fig. 5).
The measured advance rates, however, provided a range of
potential arrival times that were used to develop contingency
plans. For example, the county expended significant effort to
ensure that alternate routes were operational in advance of the
potential inundation of Highway 130, which would have cut
off several thousand residents from the rest of the island.This
effort cost about US$15 million, but will mostly be reimbursed
by FEMA (Nakaso, 2015a) thanks to a Disaster Declaration
signed by President Obama on 3 November 2014 (Moseley,
2014).Attempts were also made to time the deployment of
National Guard units and other public safety personnel and to
harden power (Fig. 3C) and water supply systems. Infrastructure
protection before the arrival of the lava flow involved
consultations by utility companies with HVO and University of
Hawai‘i at Hilo researchers and is an example of collaborations
that formed during the crisis.
Thus far, direct impacts from the June 27th lava flow have
been minimal, considering the potential destruction. Only
one house was lost, one utility pole destroyed (power was
8
sustained only minor damage and returned to service in March
2015. Indirect impacts, however, were substantial. Real estate
values plummeted as the flow advanced toward Pāhoa, and
moved away from the region (Nakaso, 2015b). Since flow
activity near Pāhoa ceased in March 2015, however, the housing
market has rebounded (Stewart, 2015).
As soon as the June 27th lava flow threat was realized,
public debate began regarding lava flow diversion. Some
residents supported intervention to protect homes and
businesses, while others strongly opposed any interference,
due to the long-term cost of diversion efforts once started,
potential for litigation, cultural sensitivities, and the low
probability of long-term success. Ultimately, diversion was not
attempted in any official capacity.
NEW INSIGHTS INTO LAVA FLOW EMPLACEMENT
The June 27th flow is significant for not only the havoc
it has caused, but also its globally unprecedented style. The
overall low effusion rate (1–2 m3/s), lack of limits on its extent
(by the ocean, for example), interaction with existing ground
cracks, and 20 km length (longest at Kīlauea in at least the
past 500 years; Clague et al., 1999) combine to make the June
27th flow the longest-lived low-effusion-rate pāhoehoe flow
known on Earth. Data collected throughout its evolution
has helped to elucidate lava flow processes, which may aid in
future crises in Hawai‘i and elsewhere.
Pressure fluctuations in the summit magma reservoir had a
significant impact on the behavior of flows that were ~40 km
downrift. Stalling of the flow front was generally preceded
by summit deflation, while rapid flow advance at the stalled
front or of upslope breakouts often followed summit inflation
(Fig. 5). Cyclic deflation-inflation (DI) events at the summit are
a common occurrence at Kīlauea (Anderson et al., 2015) and
clearly impact eruptive vigor from Pu‘u ‘Ō‘ō, where the slightly
attenuated tilt cycles are also recorded (Orr et al., 2015b).The
June 27th flow further demonstrated the importance of these
relatively minor pressure fluctuations within the summit magma
reservoir on distal lava flow activity. Summit deformation may,
in fact, provide a means of forecasting effusion rates (Patrick
et al., 2015a). Decreases in flow-front vigor were also impacted
by upslope leaks along the length of the lava tube system,
highlighting another control on flow advance.
The data that connect summit deformation to flow advance
may be useful for improving models of pāhoehoe flow hazards
by providing a means of anticipating changes in effusion and
advance rates. Lava-emplacement models, coupled with high-
resolution topography, do a good job of approximating flow
pathways (e.g., Favalli et al., 2005; Harris and Rowland, 2015),
but forecasting advance rate remains problematic, especially for
pāhoehoe flows. No forecasts of the June 27th flow advance rate
proved accurate, although they were still useful to emergency
managers for understanding the time frames of potential
impacts. Research into the physics of pāhoehoe flows—
particularly the interaction among topography, effusion rate,
lava tube formation and collapse, and flow characteristics (e.g.,
crystallinity, volatiles, temperature, and composition)—is needed
to address this limitation in current models.