Ajl. Harris et al., Automated, high temporal resolution, thermal analysis of Kilauea volcano, Hawai'i, using GOES satellite data, INT J REMOT, 22(6), 2001, pp. 945-967
Thermal data are directly available from the Geostationary Operational Envi
ronmental Satellites (GOES) every 15 minutes at existing or inexpensively i
nstalled receiving stations. This data stream is ideal for monitoring high
temperature features such as active lava flows and fires. To provide a near
-real-time hot spot monitoring tool, we have developed, tested and installe
d software to analyse GOES data on-reception and then make results availabl
e in a timely fashion via the web. Our software automatically: (1) produces
hot spot images and movies; (2) uses a thresholding procedure to generate
a hot spot map; (3) updates hot spot radiance and cloud index time series;
and (4) issues a threshold-based e-mail alert. Results are added to http://
volcano1.pgd.hawaii.edu/goes/ within similar to 12 minutes of image acquisi
tion and are updated every 15 minutes.
Analysis of GOES data acquired for effusive activity at Kilauea volcano ( H
awai'i) during 1997-98 show that short (<1 hour long) events producing 100
m long ( 10(2) to 10(3) m(2)) lava flows are detectable. This means that ti
me constraints can be placed on sudden, rapidly evolving effusive events wi
th an accuracy of <plus/minus>7.5 minutes. Changes in activity style and ex
tent can also be documented using hot spot size, intensity and shape. From
radiance time series we distinguish ( 1) tube-fed activity ( low radiance,
<10 MW m(2) m(-1)); (2) activity pauses (no radiance); ( 3) lava lake activ
ity ( low radiance, <5 MW m(2) m(-1)); ( 4) short (<3 km long) flow extensi
on (moderate radiance, 10-20 MW m(2) m(-1)); and (5) 12 km long flow extens
ion ( high radiance, 15-30 MW m(2) m(-1)).
The ability of GOES to detect short-lived effusive events, coupled with the
speed with which GOES-based hot spot information can be processed and diss
eminated, means that GOES offers a valuable additional volcano monitoring t
ool.